Home

OC-WITM-OPREF - Teledyne DALSA Inc

1. Headers include WITHOME h wMeas h Libraries WITHOME lib wObj lib WITHOME lib wMeas lib grDistMap Produce a distance map from a graphic object a0 Myj Description The grDistMap operator produces a distance map from a graphic object or vector of graphic objects The output is a float image in which each pixel contains the minimum distance to a graphic object in the input The size of the output image is specified using the width and height parameters The fill parameter determines whether rectangles circles and polygons are considered to be filled When this parameter is set to No distances are measured from the boundaries of the graphic regardless of whether the pixels are inside or outside the graphic When it is set to Yes pixels inside a graphic are set to zero If fill is set to Data a graphic is considered filled if its fillColor active field is set In some cases particularly for a large number of graphics and for polygons a distance map can be created more quickly by rasterizing the graphics with the rasterize operator and creating the distance map from the resulting binary image using the distMap operator This approach does however produce a less accurate distance map due both to rasterization and approximations made in the image distance map algorithm Parameter Information width Type int Default 100 height Type int Default 100 fill Type int Default 1 Values No 0 Yes 1
2. WITHOME lib wObj lib WITHOME lib wFit lib fitTwoLinesToPts Fit right angle or parallel lines to two sets of points C gt Description The fitTwoLinesToPts operator fits a lines to each of two input point sets using a least square fit with the constraint that the lines are either at right angles or are parallel The operator s input may be a vector of Point Fpoint PtValue Graphic Geometry or Edge objects or a single Graphic or Geometry object If the input is a vector of Graphic Geom or Edge object each object in the vector must of point type If the input is a Graphic or Geometry object it may be any type except circle and text In this case the line will be fit to the vertices treated as independent points If the input is a vector of PtValue objects the value field of the elements of the vector will be used as a weight value for a weighted least squares fit Negative weight values are ignored The operator produces two line type Graphic objects with LineO corresponding to the input point set Pts0 and Linel corresponding to the input point set Pts1 The output lines minimize the sum of the squared distances from the line to the points in the corresponding point set The lines parameter is used to specify whether the two output lines are at right angles or are parallel Parameter Information lines Type int Default 0 Values Right Angle 0 Parallel 1 Demo iGraph Operators Geometry convexHul
3. See also if if2 if4 Parameter Information cond Type int Default 0 Values False 0 True 1 Demo iGraph Operators clearObjs 117 if4 Conditional branch for 4 inputs tol biil bi LELLI Description Similar to the if operator but controlling 4 objects with the same condition See also if if2 if3 Parameter Information cond Type int Default 0 Values False 0 True 1 Demo iGraph Operators clearObjs ifCond2 Conditional branch for 2 inputs with built in test Description Similar to the ifConditional operator but controlling 2 objects with the same condition See also ifConditional ifCond3 Parameter Information condition Type int 118 Default 0 Values 0 gt 1 lt 2 3 gt 4 lt 5 Empty 6 constant Type CorString Default Demo iGraph Operators clearObjs ifCond3 Conditional branch for 3 inputs with built in test F amp d lt a z reece Description Similar to the ifConditional operator but controlling 3 objects with the same condition See also ifConditional ifCond2 Parameter Information condition Type int Default 0 Values 0 gt 1 lt 2 3 gt 4 lt 5 Empty 6 constant Type CorString Default Demo iGraph Operators clearObjs ifConditional Conditional branch with built in test Description The ifConditional operator provides conditional branching ability in an i
4. The operator will accept integer floating point input images and color images in the case optimisation parameter value is speed The output image is the same type as the input image except when filters with negative weights are applied to unsigned integer images 8 or 16 bit In this case the output image type will be a signed integer with the same number of bits as the input image type This may result in truncation of out of range values If truncation occurs relative image values can be maintained by increasing the value of the scale parameter or if the absolute image values are important the input image can be cast to a higher precision type 16 bit for 8 bit images float for 16 bit images Parameter Information kernel Type Corlmage Default 1 1 1 1 1 1 1 1 1 scale Type int Default 9 optimization Type int Default 1 Values preciseness 0 speed 1 Demo iGraph Operators Filters conv2d igr C Prototype CorOpRtn cor_conv2d CorImage In CorImage Out CorIntImage kernel int scale int optimization CorOpRtn cor_conv2d_consume CorImage In CorIntImage kernel int scale int optimization Headers include WITHOME h wFilter h Libraries WITHOME lib wObj lib WITHOME lib wFilter lib edgeDetect Create a binary or gray scale edge image Description The edgeDetect operator finds edges in images by detecting areas of high slope using Prewitt or Sobel gradient typ
5. Libraries WITHOME lib wObj lib WITHOME lib wSearch lib SmartWeb Continuous web and line scan camera specific functions 269 webClusterFeatures Consolidate clusters of features in a stream of images afl ate Description The webClusterFeatures operator groups Features from the input vector that correspond to clusters of blobs in a stream of input image The Features for each cluster are merged to produce a single Feature in which the field values are calculated over all blobs of the cluster When used with Feature vectors produced by the webGetBlobs operator the webClusterFeatures operator allows the clusters of blobs to extend across several images in a sequence of images representing vertically adjacent blocks in a larger image space A blob is assigned to a cluster when a circle or ellipse derived from the its best fit ellipse overlaps a similar ellipse from some other blob in a cluster The size of the circle or ellipse is controlled by the operator s parameters Inputs In The In input is a vector of Features produced by the getFeatures operator For this operator the Features are typically based on blobs extracted by the webGetBlobs operator Outputs Out The Out output is a vector of Features representing values calculated over the clusters found in the input Feature vector Since these Features do not represent contiguous blobs in original image they cannot be used as input to the webGetPerimet
6. Parameter Information type Type int Default 0 Values minima 0 maxima 1 vicinity Type float Default 1 201 Demo iGraph Operators Statistics HistThresh C Prototype CorOpRtn cor_histPeaks CorFpointVector Hist CorFpointVector Peaks int type float vicinity Headers include WITHOME h wStats h Libraries WITHOME lib wObj lib WITHOME lib wStats lib histogram Compute a frequency distribution a a Description The histogram operator computes a frequency distribution of values in an integer or float input image Either normal or cumulative histograms can be computed The lo and hi parameters determine the range of pixel values used in forming the histogram This range is divided into equal size bins with the number of bins specified by the nbins parameter A bin frequency is determined by incrementing the bin for each image pixel whose value falls within the bin or in the case of a cumulative histogram whose value is less than the bin maximum A normal or cumulative histogram is selected using the type parameter The operator produces a vector of Real points The first coordinate of the point is the bin boundary value and the second coordinate is the bin 202 frequency The vector may be viewed by connecting the histogram output to the graph operator for plotting If the useMask parameter is set to No the histograms are calculated using the entire image ROI If the useMask par
7. RangeLow Type float Default 0 RangeHigh Type float Default 255 Equalize Type int Default 2 Values Actual 0 User 1 Max 2 EqLow Type float Default 0 EqHigh Type float Default 255 XScale Type float Default 1 YScale Type float Default 1 Inspect Type int Default 0 Values bitwise OR X Profile 1 Y Profile 2 Magnifier 4 Track Type int Default 0 Values Mouse drags 0 Mouse moves 1 Colormap Type int Default 0 Values Grayscale 0 Pseudo color 1 Shared 2 Custom 3 ColormapName Type CorFile prom pt Show a prompt message ath Description The prompt operator displays a message in a pop up message box and may wait for a user response The message is displayed when the operator receives any object at its input The message displayed is specified by the Message string parameter The Type parameter specifies the user interaction and operator output In the Display mode no user interaction is required an integer 0 is produced immediately In the Acknowledge mode a single check box appears with the message The operator produces an integer 7 when the user clicks on the check box In the Conditional mode two check boxes appear with the message The operator produces an integer output when either of the boxes is selected If the box with the green check mark is selected the output is a 7 if the box with the red X is selected the output is a 0 Parameter Information T
8. The image will be broken into the equal size tiles If the number of pixels in the model image does not divide evenly by the number of tiles remaining pixels on the right and or bottom of the model image will be dropped colorMode The colorMode parameter controls the normalization of color images When this parameter is set to brightness the model image and target image are converted to gray scale images corresponding to the Y channel of the CIE Yuv color space and the match score is calculated using the gray scale pixel values The other color space modes are not yet implemented If the model image is not a color image the parameter value is is ignored and the color mode is automatically set to brightness Parameter Information subPixel mask tileCols tileRows Type int Default 0 Values No 0 Yes 1 Type int Default 0 Values No 0 Yes 1 Type int Default 1 Type int Default 1 colorMode Type int Default 0 Values brightness 0 fixColor 1 varyColor 2 C Prototype CorOpRtn cor_patInit2D1 CorVector ModelImages CorVector MaskImages CorVector PatModels int subPixel int mask int tileCols int tileRows int colorMode Headers include WITHOME h wSearch h Libraries WITHOME lib wObj lib WITHOME lib wSearch lib patinterpolate Sub pixel shift an image using interpolation Description The patInterpolate operator shifts the input image by non
9. name The name of the curve or curve set If the input is a list of curves you can enter muliple names E g if you are sending a vector of 3 curves then you can set the name parameter to apples oranges a complex name then the curves will be labeled accordingly on the graph If only one name is specified then each individual curve is labeled with this name plus an identification number e g curve 1 curve 2 etc Placement sub panel Placement If Fix position is checked the parameters X and Y are used to position the window Otherwise the current automatic window placement strategy is used to position the window The X horizontal position of the window on the screen if Placement has Fix position checked The Y vertical position of the window on the screen if Placement has Fix position checked width The width of the window used to display the surface plot height The height of the window used to display the surface plot Property Panel Liye Double Sings Dutmatz C Semi fe Biy O mese E Show hiogan 177 Levels If Double is chosen two threshold levels are needed If Single is chosen only one threshold level is needed Output If Semi is selected the threshold range will preserve its original pixel values pixel values which are outside the threshold range will be black If Binary is selected pixels in the threshold range will be turned white In
10. 255 showArrow Type int Default 0 Values no 0 yes 1 arrowSize Type int Default 12 arrowR Type int Default 255 Values 0 255 arrowG Type int Default 255 Values 0 255 arrowB Type int Default 0 Values 0 255 C Prototype CorOpRtn cor_plotEdges CorObj Edges CorGraphicVector PlottedEdges int r int g int b int showArrow int arrowSize int arrowk int arrowG int arrowB Headers include WITHOME h wConvert h Libraries WITHOME lib wObj lib WITHOME lib wConvert lib plotFeatures Convert Blob feature vectors to graphic descriptions Description The plotFeatures operator converts the input Blob feature object or vector of Blob features into a graphic object or vector of graphic objects for viewing with the overlayData operator The type of feature for viewing is specified by the featureType parameter and can be one of centroids ellipse axes bounding boxes perimeters or centers of bounding boxes In order to plot perimeters input Blob features must include the optional perimeter data Blob feature vectors produced by the getBlobFeatures operator with its perimeter parameter set to Yes include perimeter data Blob feature vectors produced by the getBlobFeatures operator with this parameter set to No or by the getBlobs getFeatures combination do not contain perimeter data The perimeter data can be added to these blob features by passing them through the get
11. 604 435 2587 ext 9 Email wit corecoimaging com Corporate Headquarters Coreco Imaging Inc 7075 Place Robert Joncas Suite 142 St Laurent Quebec H4M 222 Canada Tel 514 333 1301 Fax 514 333 1388 US Sales Office Coreco Imaging Inc 900 Middlesex Turnpike Building 8 Floor 2 Billerica MA 01821 USA Tel 781 275 2700 Fax 781 275 9590 307 308
12. A maximum number of iterations is specified by the maxlIterations parameter The operator will continue until it converges to a local solution that is there is no further change in mean values or the maximum number of iterations is reached As mentioned there is not generally a unique kmeans partition With inappropriate initial values the resulting partition may not be what is expected The vector outputs of the operator can be used to evaluate the quality of the solution and to help determine reasonable starting values In the simplest case the mean output from a model image can be provide the initial mean values for test images either by connecting the output directly to the meanValues parameter port with the parameter in input mode or by entering the values using the array editor The variance and number outputs produce vectors representing the variance and number of pixels for each cluster Depending on the application a cluster with a large variance or small number of pixels relative to the other clusters may indicate a problem with the partion In some cases these values could be used to modify the values in the mean vector to provide a new set of initial values Parameter Information initMeans Type int Default 1 Values array 0 auto 1 192 meanValues Type CorVector Default 0 255 numClusters Type int Default 2 maxIterations Type int Default 10000 Demo iGraph Operators Segmentation K Mean
13. Allow first and subsequent tokens to be routed differently Description The oneshot operator passes the first object it receives to the First output when in an armed initial state After an one object has flowed through the operator is in a triggered state and sends all subsequent inputs to the Subsequent port This operator is necessary to avoid some forms of igraph deadlock that can arise when using the for loop or if then else operators Using a oneshot also makes sense when you only want to do a particular igraph operation once such as defining an area of interest Parameter Information trigger Type int Default 0 Values armed 0 disarmed 1 121 Demo iGraph Operators clearObjs out Hierarchical operator output port out In Object 2 9 Out Object Description The out operator is used to associated an output port of a super node to an exit point on its sub graph The instance name of out is used as the name of the port in the super node Therefore you should choose meaningful names for each instance of the out operator Demo iGraph Operators clearObjs select5 Select 1 of 5 possible inputs for output Description Select 1 of 5 possible inputs for output Parameter Information choice Type int Demo iGraph Operators clearObjs 122 select9 Select 1 of 9 possible inputs for output wirtrirt i F Description Select 1 of 9 possible inputs for output Parameter Information
14. CorImage In CorImage Out int direction 36 Headers include WITHOME h wColor h Libraries WITHOME lib wObj lib WITHOME lib wColor lib getChannel Get a single channel from a multichannel image E 2 i Description The getChannel operator extracts a single channel from a multichannel input image The input image must be a RGB HSV or Yuv image The output is an 8 bit unsigned image The channel parameter specifies which channel is extracted from the input image For RGB HSV and Yuv images respectively channel 0 is the red hue or Y intensity channel channel is the green saturation or u channel and channel 2 is the blue value or v channel Parameter Information channel Type int Default 0 Values 0 0 1 1 2 2 Demo iGraph Operators Color Brighten C Prototype CorOpRtn cor_getChannel CorImage In CorImage Out int channel Headers include WITHOME h wColor h Libraries WITHOME lib wObj lib WITHOME lib wColor lib mergeChannels Merge three images into a multichannel image Description The mergeChannels operator combines three grayscale input images into a multichannel output images The input images must be 8 bit unsigned images For RGB HSV and Yuv images respectively ch0 is the red hue or Y intensity channel ch is the green saturation or u channel and ch2 is the blue value or v channel The output image is an RGB HSV or Yuv image as
15. CorObj A CorObj B CorObj sum Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib addNoise Add noise to an image Description The addNoise operator injects random noise into an input image thus generating a noisy output image The number of pixels to which noise is added is determined by the affectedArea parameter This parameter specifies the probability as a percentage that a pixel will be modified and corresponds on average to the number pixels changed as a percentage of the total number of pixels in the image The affected pixels are selected randomly The type of noise is selected using the noiseType parameter and its magnitude is controlled using the noiseAmount parameter When the noiseType is random a random noise offset is added to the value of the selected pixel The amount added varies uniformly from 0 noiseAmount for unsigned images or noiseAmount 2 noiseAmount 2 for signed or float images If noiseType is spike then the value of the selected pixels is set to noiseAmount If noiseType is gaussian the selected pixel values are offset by gaussian normally distributed noise with a standard deviation equal to the value of the noiseAmount parameter If the noiseAmount parameter is set to zero the offset for a selected pixel is drawn from a gaussian distribution with variance equal to the pixel value For integer image types values whi
16. CorObj B CorObj diff Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib true Generate a non zero valued integer object j Description The true operator is used to convert any object received at its input port into a non zero valued integer object This operator is useful for input to a variety of flow control operators requiring an integer as input Demo iGraph Operators Calculate CollectAverage C Prototype CorOpRtn cor_true int Out Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib unaryOp Perform a unary operation on an image using a value i Description The unaryOp operator performs unary operations when both an image A and a constant K are received at the top and bottom input ports respectively The operator supports grayscale and color input images If the input image is a grayscale image then the input constant must be a numeric value Output image image type may be specified using the outputType parameter If the input image is a color image then the input constant may be a color object of the same type as the input image or a numeric value In this case the outputType parameter must be set to default and output image will be the same type as the input image Supported operations are A K A K k A K A K logical OR amp logical AND 29 gt gt shift
17. Create a font model for use with ocrGrayRun Description The ocrGrayInit operator produces a gray scale font model for use with the ocrGrayRun operator from a prepared font The input to this operator is an OCR font usually produced by the ocrMakeF ont operator It includes a vector of images and a vector of label strings The images are the model images used to identify characters in the font The images in the vector must be the same type the images vector will be correctly constructed if the OcrFont object is produced by the ocrMakeFont operator The strings contain the labels associated with the corresponding character image The output is a vector of gray scale font models that is used by the ocrGrayRun operator as a model to identify characters C Prototype CorOpRtn cor_ocrGrayInit_1 CorOcrFont Font CorFontModelGrayVector FontModel Headers include WITHOME h wOcr h Libraries WITHOME lib wObj lib WITHOME lib wOcr lib ocrGrayRun Perform optical character recognition on a gray scale image a i Description The ocrGrayRun operator performs optical character recognition on a gray scale image using a precalculated font model The inputs to the operator are a gray scale target image a rectangle graphic or vector of rectangle graphics specifying the ROI and a font model Matches between the patterns and the image are evaluated by a statistical technique using pixel grayscale values
18. JPEG2000 CodeStream 12 width Type int Default 100 height Type int Default 100 offset Type int Default 0 rawType Type int Default 0 Values color 0 8 bit unsigned 1 8 bit signed 2 16 bit unsigned 3 16 bit signed 4 float 5 complex 6 colorFormat Type int Default 0 Values packed 0 banded 1 padding Type int Default 2 Values none 0 prepend 1 append 2 numberFiles Type int Default 0 Values 3 0 1 1 order Type int Default 1 Values RGB 0 BGR 1 C Prototype CorOpRtn cor_readImage CorImage Out CorFile filename int type int width int height 211 int offset int rawType int colorFormat int padding int numberFiles int order Headers include WITHOME h wSystem h Libraries WITHOME lib wObj lib WITHOME lib wSystem lib readObj Read a WiT object Description The readObj operator reads a WiT object from disk when any object is received at its input The name of the file to be read is specified by the filename parameter The object search path is used to search for the specified file The object file wit file may be in either the binary or text format produced by the wrObj operator Parameter Information filename Type CorFile Default sample C Prototype CorOpRtn cor_readObj CorObj Out CorFile filename Headers include WITHOME h wSystem h 212 Libraries WITHOME lib wObj lib WITHOM
19. Parameter Information zxtype Type int Default 0 Values all 0 lo gt hi 1 hi gt lo 2 contrast Type int Default 1 zxwidth Type int Default 1 197 Demo iGraph Operators Segmentation ZeroCrossing C Prototype CorOpRtn cor_zeroxX CorImage In CorImage Out int zxtype int contrast int zxwidth Headers include WITHOME h wSegment h Libraries WITHOME lib wObj lib WITHOME lib wSegment lib Statistics The Statistics library supports data reduction by the extraction or calculation of values based on the values in the input image or vector The library includes operators that extract statistics from images or vectors operators that extract image histograms operators that extract projections from images and some additional operators that process the extracted data backgnd Compute an image background value Description The backgnd operator computes an approximate background intensity by taking the average pixel value from the four corners of the input image ROI The output is a float background pixel value 198 Demo iGraph Operators Statistics BGSubtract C Prototype CorOpRtn cor_backgnd CorImage In float Background Headers include WITHOME h wStats h Libraries WITHOME lib wObj lib WITHOME lib wStats lib bound Determine extreme values in an image or vector EF D an Description The bound operator takes an input image or vector and re
20. The Indexes1 output is a vector of integers in which each element is the index number indicating which pattern model in the PatModel input vector contained the pattern that produced the corresponding match The elements of the Indexes2 integer vector indicate which pattern within that model produced the match Together these index values identify the matching pattern in the 2 D array Parameters rejectThresh The rejectThresh parameter is used to set a rejection threshold for match scores below which no matches will be considered The parameter must be set to a value from zero to one where zero indicates no correlation between the pattern and target and one indicates a perfect match If no part of the target image matches the pattern sufficiently well empty output vectors are produced acceptThresh The acceptThresh parameter specifies an acceptance threshold above which a score value is always considered a valid match If the maximum number of matches specified by the maxMatches parameter are found with match scores above the acceptThresh value the search stops immediately This parameter is used to control early termination of the search for improved speed minContrast The minContrast parameter is used to specify the minimum contrast required in the target image to produce a match The parameter specifies the minimum standard deviation of the pixel values in a possible match region relative to the standard deviation of
21. WITHOME lib wObj lib WITHOME lib wConvert lib 45 plotGeom Convert Geometry object to Graphic object il fe Description The plotGeom operator converts a Geometry object or vector of Geometry objects to a Graphic object or vector of Graphic objects that can be displayed using the overlay operator The parameters on the Color subpanel can be used to set the color of the output Graphics Parameter Information outline Type int Default 1 Values off 0 on 1 penR Type int Default 255 Values 0 255 penG Type int Default 0 Values 0 255 penB Type int Default 0 Values 0 255 fill Type int Default 0 Values off 0 on 1 fillR Type int Default 255 Values 0 255 fillG Type int Default 0 Values 0 255 fillB Type int Default 0 Values 0 255 46 Demo iGraph Operators Blobs GetBlobFeatures igr Operators Blobs GetBlobFeaturesMasked igr Operators Blobs SelectFeatures igr C Prototype CorOpRtn cor_plotGeom CorObj Geom CorObj Gfx int outline int penR int penG int penB int fill int fillR int fillG int fil1B Headers include WITHOME h wConvert h Libraries WITHOME lib wObj lib WITHOME lib wConvert lib plotLines Convert line vectors to graphic descriptions Description The plotLines operator converts a Line object or a vector of Line objects into a vector of Graphic objects which can be displayed using the overlayDa
22. WITHOME lib wObj lib WITHOME lib wMatrix lib getPOSTNET Finds and decodes POSTNET bar code Description The getPOSTNET operator accepts a binary image and one or more rectangular regions of interest ROIs then searches the ROIs for POSTNET barcodes If a barcode is found in a region it is decoded to a string Inputs Target The input image must be an unsigned 8 bit binary image with the barcode represented as dark bars on a light background Roi A single ROI may be specified by a rectangular Graphic or Geometry object Multiple ROIs can be specified by a vector of Graphic or Geometry objects Outputs Decoded If the ROI is a single object then the output is a String object If a valid barcode is found the output string contains the characters represented by the barcode If no valid barcode is found an empty string is output If the ROI is a vector of rectangles then the operator seeks for a barcode in each ROI The output is a vector of String objects If a valid barcode is found in a ROI the corresponding element of the output vector contains the decoded string If no valid barcode is found in a ROI the corresponding output vector element is the empty string Parameters noise The MaxSize parameter specifies the minimum length of the high bars in pixels Parameter Information maxSize Type int Default 40 timeout Type int Default 100 Demo iGraph Operators Matrix POSTNET igr C Pro
23. XScale YScale Inspect Track Colormap ColormapName Format 54 the upper left hand corner of the display window If Fix position is not enabled the Y parameter is ignored Type int Default 0 Type int Default 0 Type int Default 0 Type int Default 0 Type int Default 0 Values Auto 0 User 1 Type float Default 0 Type float Default 255 Type int Default 2 Values Actual 0 User 1 Max 2 Type float Default 0 Type float Default 255 Type float Default 1 Type float Default 1 Type int Default 0 Values bitwise OR X Profile 1 Y Profile 2 Magnifier 4 Type int Default 0 Values Mouse drags 0 Mouse moves Type int Default 0 Values Grayscale 0 Pseudo color 1 Shared 2 Custom 3 Type CorFile Type int Default 0 FontFamily Type CorString Default Style Type int Default 0 Values Normal 0 Italics 1 Bold 2 Bold Italics 3 Size Type int Default 11 TextRed Type int Default 0 Values 0 255 TextGreen Type int Default 0 Values 0 255 TextBlue Type int Default 0 Values 0 255 BackRed Type int Default 255 Values 0 255 BackGreen Type int Default 255 Values 0 255 BackBlue Type int Default 255 Values 0 255 edit Edit data interactively Description The edit operator allows you to interactively modify data objects If the in
24. r i si Description The zoomSize operator scales the input image to a specific image size The output image size is specified using the xsize and ysize parameters Zero size parameters have no scaling effect Images are scaled by sampling the input image at points calculated from the relative dimensions of the input and output images no interpolation is performed The output image is the same type as the input image Parameter Information xsize Type int Default 128 ysize Type int Default 128 Demo iGraph Operators Fit warp igr C Prototype CorOpRtn cor_zoomSize CorImage In CorImage Out int xsize int ysize Headers include WITHOME h wGeom h Libraries WITHOME lib wObj lib WITHOME lib wGeom lib Measurement The Measurement library supports spatial measurement in images or geometric objects The 133 library includes operators that calculate distances and angles between geometric objects and operators that produce distance maps caliper Measure the distance between two edges i FE L Description The caliper operator measures the distance between two approximately parallel edges in an image The operator searchs for a line edge near each end of a reference line The line edge is found by fitting a line to point edges found crossing a number of short lines parallel to the reference line in the neighbourhood of the ends of the reference line The distance betwe
25. 0 255 Type int Default 0 Values 0 255 Type int Default 255 Values 0 255 BackGreen Type int Default 255 Values 0 255 BackBlue Type int Default 255 Values 0 255 surface Display 3D perspective view of image surface Description The surface operator generates a 3D surface view of an image or a vector of images The Z coordinate value and the surface color grayscale value are determined from the pixel values of the input image The output of the operator is a color image as displayed by the operator SSE Ee Tete E EF Properties Main panel name Name to be used on the window title When a new object reaches a surface operator all existing data object windows are searched to look for a matching name If an existing object window has the same name then the new object is displayed in the existing window This way you can track data changes without having too many windows If no matching window name is found a new window is popped up Viewing sub panel hAngle The horizontal viewing angle Refer to the chapter Three Dimension Viewing Widgets in the Reference manual for details The vertical viewing angle Refer to the chapter Three Dimension Viewing Widgets in the Reference manual for details vAngle The distance between the eye position and the focus position Refer to the chapter Three Dimension Viewing Widgets in the Ref
26. 0 maxHeight Type int Default 0 Demo iGraph Operators Blobs GetBlobFeatures Operators Blobs GetBlobFeaturesMasked C Prototype CorOpRtn cor_getBlobFeatures_1 CorImage In CorF CorB int int eatureVector BlobFeatures lobVector Blobs type connected int perimeters int useMask CorObj mask int int int int minPixels minWidth minHeight maxPixels int maxWidth int maxHeight Headers include WITHOME h wBlob h Libraries WITHOME lib wObj lib WITHOMB lib wBlob lib getFeatures Generate Blob feature vectors Jf ae ei a p b Famas Description The getFeatures operator computes a feature set for each blob in the blob vector input The blob vector contains run length encoded blobs or contiguous non zero regions Blob vectors are produced by the getBlobFeaures operator Each feature set describes a blob by calculating its centroid location major and minor axes bounding box etc Selected blob features may be viewed using a combination of the plotFeatures and overlayData operators The Blob feature vector elments consist of the following fields Field Description Name id Unique blob identifier npixels Number of pixels in the blob area if pixels are square xmin Minimum x coordinate of blob left edge xmax Maximum x coordinate of blob right edge ymin Minimum y coordinate of blob top edge ymax Maximum y coordinate of blob bottom
27. 1 0 2 1 4 2 8 3 16 4 32 5 Type float Default 0 Type float Default 0 maxTilesOmitted Type int Default 0 maxMatches Type int Default 1 relThresh Type float Default 0 selectMode Type int Default 0 Values LocAndPat 0 Location 1 Pattern 2 xVicinity Type int Default 0 yVicinity Type int Default 0 applyPatterns Type int Default 0 Values All 0 Sample 1 Range 2 patSampling Type int Default 1 patCenter Type int Default 0 patRange Type int Default 1 C Prototype CorOpRtn cor_patRoil CorImage Target CorGraphic ROI CorPatModell PatModel CorFpointVector Locations CorFloatVector Scores CorIntVector Indexes int scale float rejectThresh float minContrast int maxTilesOmitted int maxMatches float relThresh int selectMode int xVicinity int yVicinity int applyPatterns int patSampling int patCenter int patRange Headers include WITHOME h wSearch h Libraries WITHOME lib wObj lib WITHOME lib wSearch lib patRotate Rotate a pattern image through a range of angles Description The patRotate operator creates a vector of rotated versions of the input image The input image is rotated about its center pixel in increments specified by the operator s parameters to produce the images in the output image vector This operator can be used to produce a vector of images for input into the patInit operator The resul
28. 1 Values RGB gt Bayer 0 Bayer gt RGB 1 method Type int Default 0 Values Linear 0 ColorCorrect 1 start Type int Default 0 Values GR 0 RG 1 BG 2 GB 3 Demo iGraph Operators Color Bayer C Prototype CorOpRtn cor_bayer CorImage In CorImage Out int direction int method int start Headers include WITHOME h wColor h Libraries WITHOME lib wObj lib WITHOME lib wColor lib convertColor Perform color space conversions Description The convertColor operator can be used to convert RGB images to and from HSV Yuv and gray scale images The input and output image types must correspond to the types specified by the conversion parameter Conversion to and from Gray images require or produce 8 bit unsigned images RGB gt Gray conversions produce a gray scale image that corresponds to the Y intensity channel of a Yuv image Gray gt RGB conversions produce an RGB image with equal gray values in all channels at each pixel Parameter Information conversion Type int Default 0 Values RGB gt HSV 0 HSV gt RGB 1 RGB gt Yuv 2 Yuv gt RGB 3 RGB gt Gray 4 Gray gt RGB 5 Demo iGraph Operators Color HSV Operators Color Brighten C Prototype CorOpRtn cor_convertColor CorImage In CorImage Out int conversion Headers include WITHOME h wColor h Libraries WITHOME lib wObj lib WITHOME lib wColor lib convertColorFormat Convert be
29. A positive scroll amount shifts the object downwards Negative amounts shift the object upwards In the case of the image pixels shifted off the top or bottom edges wrap around to the bottom or top respectively Parameter Information amount Type int Default 10 Demo iGraph Operators Fit polar igr C Prototype CorOpRtn cor_scroll CorObj In CorObj Out int amount Headers include WITHOME h wGeom h Libraries WITHOME lib wObj lib WITHOME lib wGeom lib shearX Shear an image along the x axis by an arbitrary angle Description The shearX operator performs a shear of the input image along the x axis by the specified angular amount in the angle parameter The angle is given in degrees where positive angles shear the image clockwise Shearing constrains an image rotation to a single degree of freedom resulting in a stretching of the image along the x axis about the given angle The range of angle is 90 90 Parameter Information angle Type float Default 10 Demo iGraph Operators Fit polar igr C Prototype CorOpRtn cor_shearxX CorImage In CorImage Out float angle Headers include WITHOME h wGeom h Libraries WITHOME lib wObj lib WITHOME lib wGeom lib shearY Shear an image along the y axis by an arbitrary angle 131 Description The shearY operator performs a shear of the input image along the y axis by the specified angular amount in the angl
30. A two level strategy is used to speed up the search for matches The operator produces the label string associated with the character found its location and a score indicating how well the character matched the model character 235 Inputs Outputs 236 Target The target image can be any gray scale WiT image an unsigned or signed eight or sixteen bit integer image or a float image The operator does not accept color or complex images ROI The ROI or ROIs are a rectangle type Graphic object or a vector of rectangle type Graphic objects The operator searches locations in the target image for which the character will be completely contained within an ROI rectangle The ROI must be large enough to contain the pattern An ROI specifying the entire image is produced by the getRect operator A single rectangle can be created using the createRect operator by specifying its corner points The entire image can be specified using the getRect operator A vector of ROIs can be created interactively using the getData operator FontModel The font model is produced by the ocrGrayInit operator Labels The operator outputs the labels of characters found in the ROIs If a single ROI is input the operator outputs the labels as a vector of Strings sorted on the x coordinate of the location of the match If a vector of ROIs in input the operator produces a vector of String vectors The elements of this vector are the label ve
31. CorOpRtn cor_adaptThresh CorImage In CorImage Out float loThresh float hiThresh int offsetMethod float loOffset float hiOffset int invert int levels int result Headers include WITHOME h wSegment h Libraries WITHOME lib wObj lib WITHOME lib wSegment lib colorThresh Thresholds multi channel images Description The colorThresh operator thresholds multi channel images The operator produces a binary output image of all pixels with values that fall within or outside a specified volume in the image s color space The operator accepts RGB HSV or Yuv images as input and produces an 8 bit unsigned image as output The type of color space volume is specified using the type parameter When this parameter is set to Level the color space volume is the region in which the pixel value in each channel is greater than the corresponding value among the ch0 Low ch1 Low and ch2 Low parameters When the type parameter is set to Box the volume is a rectangular box with axes parallel to the color space coordinate axes when set to Ellipsoid the volume is an ellipsoid with axes parallel to the color space coordinate axes and when set to Cylinder the volume is an elliptical cylinder with the cylinder axis parallel to the channel 0 axis and the ellipse axes parallel to the channel 1 and channel 2 coordinate axes For these types the size of the volume is specified by the ch0 Low ch0 High ch1 Low chil Hig
32. Libraries WITHOME lib wObj lib WITHOME lib wFilter lib sobel Sobel edge enhancement a ib F Description The sobel operator enhances the horizontal or vertical edges of the image by performing a convolution of the input image with a Sobel horizontal or vertical kernel The direction parameter selects between the horizontal or vertical kernel types These kernel types are equivalent to using conv2d with the following kernel values and a scale of 8 T ae T Te 0 2 0 0 0 2 02 2 1 P oser The optimization parameter defines whether the fast or precise algorithm will be used The operator accepts integer and float type input images if optimization is precise Only integer input images are supported for fast The output image type is the signed version of the input image type Parameter Information direction Type int Default 0 Values horizontal 0 vertical 1 optimization Type int Default 1 Values preciseness 0 speed 1 Demo iGraph Operators Filter Entropy Operators Filter V ariance C Prototype CorOpRtn cor_sobel CorImage In CorImage Out int direction int optimization CorOpRtn cor_sobel_consume CorImage In int direction int optimization Headers include WITHOME h wEFilter h Libraries WITHOME lib wObj lib WITHOME lib wFilter lib thinEdges Reduces the width of the thick edges in grayscale images Dea Description The thinEdges operator r
33. Morphological gradient Description The morphGradient operator enhances edges in an image by subtracting the erosion of the input image from the dilation of the input image If the type parameter is set to GrayScale the pixel values in the output image are the difference between the neighbourhood maximum and the neighbourhood minimum pixel values in the input In the GrayScale mode the input image may be a gray scale or RGB image The output image type is unsigned 8 bit for unsigned or signed 8 bit inputs signed 16 bit for unsigned or signed inputs saturation may occur and the same type as the input for float or RGB inputs If the type parameter is set to Binary pixels in the output image are non zero if they are non zero in the binary dilation of the input image and zero in the binary erosion of the input image The input image may be an unsigned 8 bit unsigned 16 bit image or RGB image The output image is the same type as the input image For unsigned or signed 8 bit and float input images the optimization parameter may be set either to preciseness or to speed For unsigned or signed 16 bit images the optimization parameter must be set to preciseness and for RGB images the optimization parameter must be set to speed The neighbourhood used for the dilation and erosion operations is specified by the kernelType width height and kernel parameters as described for the erode and dilate operators Parameter Information
34. Operators Geometry convexHullL igr C Prototype CorOpRtn cor_gCentroid CorImage In CorObj ROI CorObj Centroid Headers include WITHOME h wFit h Libraries WITHOME lib wObj lib WITHOME lib wFit lib gFitContoursRoi Find straight lines within a ROI P Description Find straight lines within a ROI Parameter Information thresh Type float 104 startAngle Type int Default 128 endAngle Type int Default 127 range Type int Default 0 degrees Type int Default 0 Values 360 0 256 1 Demo iGraph Operators Geometry convexHull igr C Prototype CorOpRtn cor_gFitContoursRoi CorImage In CorObj ROI CorGeomVector Contours CorFloatVector intensity float thresh int startAngle int endAngle int range int degrees Headers include WITHOME h wFit h Libraries WITHOME lib wObj lib WITHOME lib wFit lib gFitLine Least square line fit to pixel values in ROI Description The gFitLine operator finds the line which gives the best least squares fit to the pixels of a region of interest ROI weighted by the values in the input image The input ROI can be either a graphic object or by a binary image Graphic object ROIs are rasterized and all pixels in the input image corresponding to the rasterized pixels including boundary pixels are used in the calculation of the centroid When a binary image ROI is used all pixels in the input image that
35. Out int amount Headers include WITHOME h wGeom h Libraries WITHOME lib wObj lib WITHOME lib wGeom lib polar Transform image to polar representation Description The polar operator can transform an image to and from polar or log polar space The direction parameter determines whether the input image is transformed from polar to rectangular P gt R or from rectangular to polar R gt P The polar coordinate space is centred at the point supplied at the lower input which can be a Graphic Geometry Point or Real point object The type parameter determines whether the space is polar or log polar When transforming from rectangular to polar the transformation is applied over the radius range specified in pixels by the minRadius and maxRadius parameters and over the angular range specified in degrees by the minAngle and maxAngle parameters The angles are calculated in degrees clockwise from the positive x axis The number of pixels in the output image is specified with the rBins and aBins parameters for the radial x and angular y dimensions Values outside the input image are set to zero When transforming from polar to rectangular the input image is transformed into the region defined by the minRadius maxRadius minAngle and maxAngle parameters When transforming a polar image created by another instance of the polar operator back into rectangular coordinates these parameters and the type
36. Parameter Information dir Type int Default 0 Values forward 0 reverse 1 Demo iGraph Operators Transforms Transforms C Prototype CorOpRtn cor_dct CorImage In CorImage Out int dir Headers include WITHOME h wXform h Libraries WITHOME lib wObj lib WITHOME lib wXform lib fft Fast Fourier transform an image Description The fft operator computes the Fast Fourier transform of the input image The input image may be a gray scale integer or float or complex image A forward or reverse FFT is specified by the dir parameter The output parameter is used to specify whether the output image is in the default format or in a user specified format In the Default mode the forward transform produces a complex image and the reverse transform produces the real part of the reverse transformed image as a float image This is required for backward compatability In the User mode the outType parameter is used to specify the output image format When outType is set to Complex the transform is output in the WiT complex image format The other settings produce a float image containing the Real part Imaginary part Magnitude or Phase of the transformed image The size of the output image is equal to that of the input image regardless of the input image ROI size rounded up to the nearest power of two in each dimension Only pixels in the input image ROI are used in the FFT calculation The forward FF
37. RangeHigh Equalize EqLow EqHigh Colormap Type float Default 1 Type int Default 1 Type int Default 1 Type CorString Default 255 0 0 Type int Default 0 Values bitwise OR Fix position 1 Type int Default 0 Type int Default 0 Type int Default 500 Type int Default 500 Type int Default 0 Values Auto 0 User 1 Type float Default 0 Type float Default 255 Type int Default 0 Values Auto 0 User 1 Type float Default 0 Type float Default 255 Type int Default 0 Values Grayscale 0 Pseudo color 1 Custom 2 ColormapName Type CorFile terrain Display 3D terrain 69 Pe ae Description The terrain operator generates a 3D surface terrain image based on values in the two input images The elevation Z coordinate value of the terrain is determined from the elevation input image The surface color grayscale value is determined from the pixel values in the image input image When the same image is applied to both inputs the operation of the terrain operator is identical to that of the surface operator Refer to the surface operator for details Parameter Information name Type CorString Default terrain hAngle Type float Default 45 vAngle Type float Default 45 zoom Type int Default 65 scale Type float Default 1 grid Type int Default 1 wireframe Type int Default 1 wireColor Type CorString Defau
38. Setting this parameter to a value greater than zero provides some tolerance to partially occluded or damaged target regions localSearch The localSearch parameter specifies the method used to search the local neighbourhoods of candidates found in the low resolution search When the parameter is set to Fast a binary hill climbing method is used The hill climbing method can get stuck in local a maximum in particular when the target and pattern model images contain fine regular patterns To prevent this occuring the parameter can be set to All in which case an exhaustive search of the neighbourhood is performed The exhaustive search is slower for large candidate search scale factors subPixel The subPixel parameter allows the match locations to be refined to subpixel precision The pattern model must have been prepared using the subpixel mode of the patInit operator resolution The resolution parameter specifies the resolution of the subpixel refinement of the match locations The resolution can be specified in powers of two from 2 half pixel to 64 1 64 pixel This parameter is only used when the subPixel parameter is set to Yes The Candidates Sub Panel 249 250 The patFind operator uses a two level seach strategy to speed up the matching process First a lower resolution search is performed over the entire ROI to find candidate matches The candidate matches are then refined at full and possibly subpixel
39. The 16 operator is similar to the collector operator except it produces the average or sum of images received rather than a vector of images The operator will accept integer or float gray scale images or color images All images in the sequence of input images must be the same type and must have the same ROI The output image is the same size as the input image In the Average mode the output image is the same type as the input image In the Sum mode eight bit input images produce 16 bit output images while 16 bit integer and float input images produce float output images Parameter Information number Type int Default 0 output Type int Default 0 Values Average 0 Sum 1 Demo iGraph Operators Calculate CollectAverage C Prototype CorOpRtn cor_collectAverage CorImage In CorObj Control CorImage Avg int number int output Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib comparelmages Compare pixel values in two images ehg Description The compareImages operator performs a pixel by pixel comparison of the values in two input images to produce a binary image The A and B input images may be any scalar image type uint8 int8 uintl6 int16 or float32 They do not have to be the same type The condition parameter specifies the test applied to values of corresponding pixels in images A and B The tests available are A lt B A
40. The relationship of pixels in the transformed image to lines in the reverse transformed image is the inverse of that described for the forward transform The startAngle parameter specifies the angle associated with the first column of the input image The angleIncr parameter controls the angle increment associated with each pixel in the x direction The accumulate parameter controls how the output pixel values are assigned In the max mode each output pixel is set to the maximum value of all associated pixels from the input image The resulting image contains lines whose value gives an indication of the number of active pixels that were on that line in the original image In the sum mode each output pixel is assigned the sum of all associated pixels from the input image This produces a scaled blurred reconstruction of the original image In the incr mode the output pixels are incremented for each associated pixel in the input image The pixel value gives an indication of the number of lines in the original image points in the transformed image that cross the pixel This is most useful when applied to binarized transformed images in which non zero pixels are sparse such as might be produced by thresholding the transformed image The inverse transformed image will always be larger than the original image with both width and height of the image approximately equal to the diagonal length of the original It can be resized to the origina
41. WITHOME lib wVideo lib acqCounterReset Reset acquisition event counter yori Description The first time the acqCounterReset operator executes it initializes the hardware acquisition event counter if available and resets the count to zero Subsequent executions simply reset the count to zero This operator must be executed at least once before executing acqCounterRead C Prototype CorOpRtn acqCounterReset Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqDisplay Draw an image on the video window Description Draw an image on the video window C Prototype CorOpRtn acqDisplay CorImage In Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqExposure Set the exposure duration of a camera s electronic shutter Description The acqExposure operator is used to set the amount of exposure required by the camera during a frame grab by the acquire operator The duration parameter specifies the exposure in units particular to the frame grabber usually either milliseconds or microseconds Parameter Information duration Type uint Default 16667 C Prototype CorOpRtn acqExposure uint duration Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqFrameTrigger Control a frame grabber s frame trigger mode a G Descrip
42. choice Type int Demo iGraph Operators clearObjs sequencer Output elements of a vector in sequence Description The sequencer operator dispatches the elements of a vector object received at its input port one by one to the Element output port along with an end condition flag from the Control output port When the last element is sent the end condition flag sent is set to true otherwise it is false The condition flag can be used by the collector operator see also collector Parameter Information count Type int Default 0 Demo iGraph Operators clearObjs start Start an igraph Description The start operator produces a sync object It is typically used as the starting point of an igraph because most operators require at least one input Demo iGraph Operators clearObjs stopDisable Do not allow user to stop igraph La E zg Description Do not allow user to stop igraph Demo iGraph Operators clearObjs stopEnable Enable user to stop igraph Ena agi Description Enable user to stop igraph Demo iGraph Operators clearObjs vgate Gate operator in vertical orientation H Description The vgate operator allows objects to travel from input to output if the Gate input is present otherwise the input object is blocked Both the Gate input and input object must be present before the vgate operator executes This operator is the same as the gate but the input and outputs are positio
43. decreases local contrast enhancement The operator accepts integer and float type input images An unsigned integer input image produces a signed integer output image with the same number of bits as the input image In all other cases the output image is the same type as the input image Parameter Information width Type int Default 3 Demo iGraph Operators Segmentation ZeroCrossing C Prototype CorOpRtn cor_hipass1d CorImage In CorImage Out int width CorOpRtn cor_hipassld_consume CorImage In int width 85 Headers include WITHOME h wFilter h Libraries WITHOME lib wObj lib WITHOME lib wFilter lib hipass2d Two dimensional high frequency enhancing filter Description The hipass2d operator applies high pass filtering or sharpening to the input image to enhance local contrast and suppress slow changing values This technique is helpful in enhancing image detail and edges and in eliminating gradually changing global effects such as lighting variations from camera images First a smoothed image is calculated over the image as described for the lopass1d operator The smoothed image is then subtracted from the original image to give the high pass filtered image The size of the kernel is specified by the width and height parameters Increasing the kernel size lowers the cut off frequency of the filter and decreases local contrast enhancement This operator could be executed in two mode
44. gray scale 1 threshold Type int Default 50 Values 0 100 invert Type int Default 1 Values No 0 Yes 1 filter Type int Default 0 Values None 0 2x2 1 3x3 2 5x5 3 7x7 4 charSet Type int Default 1 Values All 0 Alpha numeric 1 Digits 2 C Prototype CorOpRtn cor_ocrMakeSemiFont CorOcrFont Out int style float width float height float clearZone float strokeWidth int outType int threshold int invert int filter int charSet Headers include WITHOME h wOcr h Libraries WITHOME lib wObj lib WITHOME lib wOcr lib ocrPerimlnit Create a font model for use with ocrPerimRun Description The ocrPerimInit operator produces a binary font model for use with the ocrPerimRun operator from a prepared font The input to this operator is an OcrFont type object The OcrFont object is usually produced by the ocrMakeF ont operator It includes a vector of images and a vector of strings The images are the model images used to identify characters in the font The strings contain the labels associated with the corresponding character image The output is a FontModel object that is used by the ocrPerimRun operator as a model to identify characters It is recommended that the FontModel object only be used as internally to igraphs or programs and that fonts are stored externally in the form of the OcrFont object Parameter Information samples Type int Default 3
45. h wCalibration h Libraries WITHOME lib wObj lib WITHOME lib wCalibration lib calibrateDefault Create default camera calibration transformation ETTE ra ie Description The calibrateDefault operator creates a calibration object for a trivial one to one transformation See calibrateCamera operator for more interesting non trivial transformations calibrateDefault is useful for creating a transformation object for operators that require it but when no transformation is necessary width and height parameteres should be the same as the size of all images that will the calibration will be applied to Parameter Information width Type int Default 640 height Type int Default 480 interpolation Type int Default 0 Values perspective 0 bilinear 1 origin Type int Default 0 31 Values none 0 bottomLeft 1 topLeft 2 topRight 3 bottomRight 4 C Prototype CorOpRtn cor_calibrateDefault CorCalibrateInfo calibrateInfo int width int height int interpolation int origin Headers include WITHOME h wCalibration h Libraries WITHOME lib wObj lib WITHOME lib wCalibration lib calibrateRatio Calculate scale ratio between two camera calibrations af c Lt Description The calibrateRatio operator computes the scale ratio between two calibration transformations It is useful when some graphic or geometric objects have been created using an old transformation but
46. int Default 0 C Prototype CorOpRtn cor_getRow CorImage In CorVector Rowvec int row 177 Headers include WITHOME h wPixel h Libraries WITHOME lib wObj lib WITHOME lib wPixel lib getVector Get pixel values along an arbitrary line segment ml Description The getVector operator retrieves the vector of pixel data from the input image corresponding to the vector defined by two end points specified by the start and end parameters The pixels are selected by rasterizing the line connecting the two end points and consequently do not represent evenly spaced samples Evenly spaced samples can be obtained using the getPath operator Parameter Information start Type CorPoint Default 0 0 end Type CorPoint Default 0 0 C Prototype CorOpRtn cor_getVector CorImage In CorVector Profile CorPoint start CorPoint end Headers include WITHOME h wPixel h 178 Libraries WITHOME lib wObj lib WITHOME lib wPixel lib insert Overlay an image on a larger image DHO Description The insert operator overlays a subimage or subimages on larger images in specified locations The main use of this operator is to insert subimages that have been extracted using the extract operator back into the original image The graphic or vector of graphics used to extract the subimages can be used to insert the subimages A subimage or vector of subimages is input at the
47. is required this parameter is usually used as an input parameter When a zero is received the blobs are built up as usual When a one is received all blobs in the current image are output even if they reach the bottom of the image and after processing for the current image is complete the starting line number is reset to zero The webLineCounter operator can be used to provide the required control signals When no reset function is required this parameter is set to No In this mode the line count will continually increase as images are received until the graph stops Blobs which reach the bottom of the last image processed will not be output by the operator resetBlobId The blob id number can be set to zero with the resetBlobId parameter This will set the next blob id to zero after processing the current image This parameter is usually set to No or used as an input parameter It may be tied to the reset parameter when both are input parameters in which case a new set of id numbers will be started every time the operator is reset Parameter Information type Type int Default 0 Values binary 0 labeled 1 connected Type int Default 1 Values 4 0 8 1 maxBlobHeight Type int reset Default 0 Type int Default 0 Values No 0 Yes 1 273 resetBlobld Type int Default 0 Values No 0 Yes 1 minPixels Type int Default 0 minWidth Type int Default 0 minHeight Type int Default 0 C
48. me X meena Format TFF o Tag Image File Format Image File Format Microsoft Windows Device a e s Ee Bitmap Format JPEG PEG Joint Joint Photographers Expert Group Expert Group WiT WiT object format images only Raw raw data of known size and type possibly with a header of known hemo IPEG2000 JP2 JP2 JPEG 2000 S 2000 JPEG2000 JPEG 2000 CodeStream CodeStream Besides these image format types you can choose Auto to let the operator detect the image format type where possible Raw images cannot be detected using the Auto setting For PCX formats readImage converts an image to a WiT RGB image if the image file contains color otherwise a gray scale image is converted to an 8 bit gray scale WiT image For PBM format readImage supports bit mapped PBM gray scale PGM and color PPM types Additionally the operator interprets both ASCII text and raw formats For SUN format both gray scale and color images both pseudo or true color are supported For XBM format the operator ignores the hotspot location in an XBM file and creates an 8 bit gray scale WiT image Color is not supported For TIFF format readImage supports gray scale palette and RGB types For BMP format readImage supports 1 4 8 and 24 bit image types readImage converts an image to a WiT RGB image if the image file contains color otherwise an 8 bit gray scale WiT image is produced For the JPEG format the operator pro
49. rather than A B Parameter Information expression Type CorString Default A userConstant Type double Default 0 type Type int Default 0 Values default 0 char 1 uchar 2 short 3 ushort 4 int 5 uint 6 float 7 double 8 15 conversion Type int Default 0 Values round 0 floor 1 ceiling 2 C Prototype CorOpRtn cor_calc CorObj A CorObj B CorObj Out CorString expression double userConstant int type int conversion Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib collectAverage Produce the average or sum of a sequence of images oxy 0 Description The collectAverage operator collects a number of images and forms a single average or sum image from them The type parameter value determines whether an average or sum is formed The number of images collected is controlled by the number parameter If the number parameter is set to 0 images are accumulated with no output produced while integer 0 objects are received at the Control input When a non zero integer object is received at the Control input the input image is added to the accumulated image and an output image is produced The operator is cleared and a new sum or average is started when the next image is received If number is set to a non zero value the parameter value specifies the number of images that will be collected to form the output image
50. speed 1 Demo iGraph Operators Morphology hitormiss igr C Prototype CorOpRtn cor_rank_1 CorImage In CorImage Out CorIntImage kernel int rank int optimization CorOpRtn cor_rank_1_consume CorImage In CorIntImage kernel int rank int optimization Headers include WITHOME h wMorpho h Libraries WITHOME lib wObj lib WITHOME lib wMorpho lib skeleton Thin binary image with skeleton constraint b Description The skeleton operator is used to remove pixels from objects in a binary input image until a single pixel 151 wide skeleton is reached On each pass objects in the image are eroded by a one pixel wide boundary provided the pixels are not part of a single pixel wide skeleton After each execution of this operator an output image is generated at the top output port along with a pixelsRemoved value at the bottom output port This value represents the number of pixels removed from the image and is an integer object When the pixelsRemoved value reaches zero the objects have reduced to a single pixel wide skeleton The numPasses parameter controls the number of erosion operations applied to the input image If this value is set to zero the image will be eroded until no more pixels are removed and the skeleton image is produced Single pixel wide skeletons can also be produced by applying the skeleton operator iteratively using the if operator controlled by the value of pixel
51. the threshold based on the other parameters as described below The selected points are output as a vector of graphic point objects The numToMatch applies a limit to the number of outputs collected If more than numToMatch matches are found the selectMatches parameter determines if the first matches found or the best matches found are reported If the first matches are specified the operator stops searching when numToMatch values above the threshold are found If the best matches are required the entire set of pixels specified by the input graphic object and the searchRange parameter in the case of points is searched The vicinity parameter is used to compare matched objects with similar position If a new match is found which is within vicinity pixels in x and y of a previously found match point then only the better of the match points is recorded Parameter Information numToMatch Type int Default 5 selectMatches Type int Default 0 Values First 0 Best 1 vicinity Type int Default 5 searchRange Type int Default 5 threshold Type float Default 0 5 find Type int Default 0 Values maxima 0 minima 1 color Type int Default 0 Values default 0 select 1 red Type int Default 255 Values 0 255 green Type int Default 0 Values 0 255 blue Type int Default 0 Values 0 255 Demo iGraph Operators Find NXCorr Operators Find SoD Operators Find X Corr C Prototype CorOpRtn c
52. viewed by the operator The current displayed image can be saved to disk as a WiT image at any time using the Save button on the image properties panel Any number of images can be saved from each invocation of the volume operator by changing the name of the stored image for each save operation Properties Main panel name Name to be used on the window title When a new object reaches a volume operator all existing data object windows are searched to look for a matching name If an existing object window has the same name then the new object is displayed in the existing window This way you can track data changes without having too many windows If no matching window name is found a new window is popped up Placement sub panel Placement If Fix position is checked the parameters X and Y are used to position the window Otherwise the current automatic window placement strategy is used to position the window If Max size is checked then the parameters MaxW and MaxH are used to limit the maximum size of the window If the object requires a window larger than the maximum size then scroll bars will appear X The X horizontal position of the window on the screen if Placement has Fix position checked Y The Y vertical position of the window on the screen if Placement has Fix position checked Max W The maximum width of the window if Placement has Max size checked MaxH The maximum height
53. written as quoted character strings all other objects are written in the string format produced by the cast operator without quotes Vectors are written as strings of comma separated values on a single line If the elements of the vector are of String type the individual elements are placed in quotation marks The subvectors of a vector of vectors are printed in the same way with each subvector on a single line The name of the file to which the string is written is specified by the filename parameter The current directory is used unless an absolute path name is given The fileMode parameter may be set to write mode in which the new value overwrites any existing data in the file or append mode in which the new value is added to the end of the existing data The text files produced by wrText cannot be read back into WiT Files that can be read into WiT can be written in either binary or text format by the wrObj operator Parameter Information filename Type CorFile Default test txt format Type int Default 0 Values unlabelled 0 labelled 1 fileMode Type int Default 1 Values write 0 append 1 C Prototype CorOpRtn cor_wrText CorObj In CorFile filename 214 int format int fileMode Headers include WITHOME h wSystem h Libraries WITHOME lib wObj lib WITHOME lib wSystem lib writelmage Write WiT image in a standard image file format Description The writeImage oper
54. 1 1 1 1 1 1 1 1 1 optimization Type int Default 1 Values preciseness 0 speed 1 Demo iGraph Operators Morphology skeleton igr C Prototype CorOpRtn cor_thin CorImage In CorImage Out CorIntImage structElem int optimization CorOpRtn cor_thin_consume CorImage In CorIntImage structElem int optimization Headers include WITHOME h wMorpho h Libraries WITHOME lib wObj lib WITHOME lib wMorpho lib tophat Tophat filter Description The tophat operator finds pixels which exceed a threshold relative to specified pixels in their neighbourhood The neighbourhood pixels that are considered are specified by the kernel parameter Neighbourhood pixels that correspond to 1 s in the kernel are active while those corresponding to 0 s in the kernel are ignored 153 A pixel in the output image is set if the value of the corresponding pixel in the input image is greater than or equal to the maximum of the active neighbourhood pixels plus the threshold for positive thresholds or less than or equal to the minimum of the active neighbourhood pixels plus the threshold for negative thresholds All other pixels in the output image are set to zero If the overlay parameter is turned on the set pixels are assigned the value of the corresponding pixel in the input image otherwise the set pixels are assigned the maximum image value The tophat operator requires an 8 or 16 bit unsigned image as its
55. B RGB images Complex images are converted by calculating the magnitudes of the complex values Other image types are converted to complex by assigning the pixel gray levels the luminance in the case of color images to the real part of the corresponding complex pixel the imaginary parts are all set to zero When float image types are converted to an integer type pixel values are rounded to the nearest integer value The matchSize operator can be used to prepare images for processing with the fastAluOp which requires that images be the same size and offset Parameter Information sizeOp Type int Default 0 Values clip B 0 expand A 1 clip A and B 2 center B on A 3 offset Type CorPoint Default 0 0 outputType Type int Default 0 Values default 0 8 bit unsigned 1 8 bit signed 2 16 bit unsigned 3 16 bit signed 4 float 5 complex 6 color 7 Demo iGraph Operators Fit warp igr C Prototype CorOpRtn cor_matchSize CorImage A CorImage B CorImage Aout CorImage Bout int sizeOp CorPoint offset int outputType Headers include WITHOME h wGeom h Libraries WITHOME lib wObj lib WITHOME lib wGeom lib move Rotate an object by an arbitrary angle Description The move operator performs an arbitrary rotation and transformation on its input object Rotation is executed first angle parameter in degrees counterclockwise followed by translation shift parameter
56. B The expression field allows the user to specify a C style mathematical equation which will determine the output image The top input image can be referenced in this equation using the variable A while the bottom image is given using B The result for each pixel can be affected by that pixel s particular column and row using the X and Y variables respectively For example the expression Y gt 5 A B will produce an output image where the first 5 rows are identical to the first five rows of B Otherwise the output image corresponds to A For more on the expression syntax and use see calc This operator is faster than aluOp but has some limitations on manipulation of the image ROI and does not allow the specification of the output image type Relative offset can be set and image size matched using the matchSize operator Parameter Information aluOp Type int Default 0 Values 0 1 2 3 OR 4 AND 5 XOR 6 src 7 overlay 8 expr 9 abs A B 10 min 11 max 12 overlayMask Type CorHex Default Oxffff expression Type CorString Default A Demo iGraph Operators Calculate CollectAverage C Prototype CorOpRtn cor_fastAluOp CorImage A CorImage B CorImage Out int aluOp CorHex overlayMask CorString expression Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib fastUnaryOp Faster unaryOp Description The fastUn
57. Blob features is used they must include the Blob perimeter data The Blob feature vector can be created by applying the getBlobs getFeatures and getPerimeter operators to a binary image 242 FontModel The precalculated font model is created using the ocrPerimInit operator The font model should be based on a clean representative example of the font The font model contains the matching features associated with each model character as well as a label that represents the character More than one entry in the font model may have the same label Such cases represent variations in the shape of a character The font model can be modified using the ocrModify operator Outputs Labels The operator outputs the labels of characters that match blobs in the input image If a single ROI is input the operator outputs the labels as a vector of Strings sorted on the x coordinate of the location of the match If a vector of ROIs in input the operator produces a vector of String vectors The elements of this vector are the label vectors of the corresponding ROI in the input ROI vector Scores The operator outputs a score of for each character character found that indicates the quality of the match with the model character Scores range from 0 to 1 with 1 indicating a perfect match to the model character The scores are output as a vector of float values or a vector of float vectors depending on whether the input ROI is a single ROI or a vector
58. CorObj ROI CorObj Decoded int noise Headers include WITHOME h wMatrix h Libraries WITHOME lib wObj lib WITHOME lib wMatrix lib getCode39 Finds and decodes Code 39 bar code Description The getCode39 operator accepts a binary or grayscale image and one or more rectangular 224 regions of interest ROIs then searches the ROIs for Code39 barcodes If a barcode is found in a region it is decoded to a string Inputs Target The input image must be an unsigned 8 bit grayscale image with the barcode represented as dark bars on a light background Roi A single ROI may be specified by a rectangular Graphic or Geometry object Multiple ROIs can be specified by a vector of Graphic or Geometry objects Outputs Decoded If the ROI is a single object then the output is a String object If a valid barcode is found the output string contains the characters represented by the barcode If no valid barcode is found an empty string is output If the ROI is a vector of rectangles then the operator seeks for a barcode in each ROI The output is a vector of String objects If a valid barcode is found in a ROI the corresponding element of the output vector contains the decoded string If no valid barcode is found in a ROI the corresponding output vector element is the empty string Parameters noise The noise parameter specifies the minimum low to high or high to low change in pixel value necessary to constitu
59. Data 2 Demo iGraph Operators Measurement CircleCenters Operators Segmentation Watershed C Prototype CorOpRtn cor_grDistMap CorObj In CorImage Out int width int height int fill Headers include WITHOME h wMeas h Libraries WITHOME lib wObj lib WITHOME lib wMeas lib plotAngle Compute and plot the angles between points in a vector PE aad Description The plotAngle operator calculates the angles between adjacent points in a vector with respect to a specified vertex point and creates a vector of graphic objects which allow the angles to be displayed on an image The vector of radial points is received at the top input the vertex at the bottom input The inputs may be a Points Real points point type Graphics point type Geometry objects or Edge objects with a point type geometry If the input vector contains N points N J angles are calculated for the lines connecting the vertex to points 0 and 1 to 1 and 2 to N 2 and N J A vector of Graphics is created which includes the vertex and radial points and lines connecting them and the angle value in degrees These may be displayed on an image using the overlayData operator The color of the graphical objects can be selected from a number of color combinations with the color parameter The points and text are printed in the main color the connecting lines are printed in the secondary color The font fontSize and fontStyle
60. Description The complxMult operator performs pixel by pixel multiplication of two complex images or of a complex image and a real image The A port accepts a complex image and the B input port accepts a real integer or float image or a second complex image The output complex image is the product of the two input images The operator is usually used for processing complex images produced by a forward fft operation Complex images may also be produced from pairs of float or integer images representing the real and imaginary components using the complexMerge operator The output image is the same size as the A input image The output image ROI is the intersection of the ROIs of the two input images Image values outside the ROI are undefined 17 Demo iGraph Operators Transforms FourierFilter C Prototype CorOpRtn cor_comp1xMult CorImage A CorImage B CorImage Product Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib conditional Compare numbers or strings and output a boolean result Description The conditional operator performs conditional testing on two inputs The user selects the condition operation using the condition parameter Currently the following conditionals are supported A V Il AV iow The inputs must be one of the basic WiT types i e an integer float character or string The output type is an integer and is set to 1
61. Description The patRotate operator creates a vector of image vectors representing scaled and rotated versions of the input image The input image is scaled in increments specified by the operator s parameters Each scaled image is rotated as specified by the 267 parameters to produce a vector of images These vectors of rotated images are collected to produce the output vector of image vectors This operator can be used to produce a vector of image vectors for input into the patInit2D operator The resulting pattern model vector can then be used to search for scaled rotated patterns using the patFind2D operator Inputs In The input image can be a gray scale or color image complex images are not currently supported by this operator Outputs Transformed The input image is first scaled then each scaled image is rotated to produce a vector of images These vectors of rotated images are collected to produce the output vector of image vectors The number of scaled image vectors and their scale factors the number of rotated images in each vector and their rotation angles and the image sizes are controlled by the parameters as described below The images in the vectors are of the same type as the input image Params The Params output is a four element float vector that contains the transformation parameters applied to the images in output vector The first element is the minimum scale factor the second is the scale increment th
62. E Description The makeSeq operator produces a sequence of numbers as a WiT vector by parsing a C expression representing the desired mapping function from the vector index values to the vector element values The number of elements in the sequence vector is specified by the size parameter The element values are calculated are as a function of the array index values from 0 to size 1 as defined by the parameter expression The symbol A is used to represent the index value in the expression The symbol K is optionally used in the expression to represent a constant value that is specified in the userConstant parameter For more on the syntax and usage of expression see calc The type of the elements in the output vector is specified by the type parameter Calculations are performed at double precision and cast to the required output type after the calculations are completed The conversion parameter can be used to specify whether the values are rounded to the nearest round nearest lower floor or nearest greater ceil integer value when one of the integer types is specified The byte ubyte short and ushort output values outside the range of the type are truncated to the minimum or maximum range values Parameter Information expression Type CorString Default A size Type int Default 255 userConstant Type double Default 0 type Type int Default 4 Values byte 0 ubyte 1 short 2 ushort 3 int 4 ui
63. Fit warp igr C Prototype CorOpRtn cor_warp CorObj In CorFloatVector Transform CorObj Out Headers include WITHOME h wGeom h Libraries WITHOME lib wObj lib WITHOME lib wGeom lib zoomScale Zoom an object by scaling x and y ag Description The zoomScale operator performs arbitrary scaling of the input object The x and y scaling factors specified using the xscale and yscale parameters respectively Images are scaled by sampling the input image at points calculated from the scaling factors no interpolation is performed The output image is the same type the input image Graphical objects are scaled with respect to the origin that is the upper left corner of the image space so a graphical object scaled by a number greater than one will become larger and move away from the upper left corner Scaled graphical objects will retain their relative position when overlayed on similarly scaled images The operator accepts vectors of graphical objects such as those produced by getData or single objects Parameter Information xscale Type float Default 0 5 yscale Type float Default 0 5 Demo iGraph Operators Fit warp igr C Prototype CorOpRtn cor_zoomScale CorObj In CorObj Out float xscale float yscale Headers include WITHOME h wGeom h Libraries WITHOME lib wObj lib WITHOME lib wGeom lib zoomSize Zoom an image by setting width and height i gt
64. G i and the filtered version of G i Notationally G 1 input image G i reduce G i 1 for i gt 1 L i G i gauss G i The filter parameter selects the type of gaussian filter to be applied The effect of each filter is described in the WiT filter library during the discussion of the gauss operator The minsize parameter defines the minimum size that an image G i can be before a reduce operation will not be performed and pyramid generation will terminate Parameter Information filter Type int Default 0 Values 2x2 0 3x3 1 5x5 2 7x7 3 minsize Type int Default 8 C Prototype CorOpRtn cor_fsd CorImage In CorImageVector Gaussian CorImageVector Laplacian int filter int minsize Headers include WITHOME h wPyramid h Libraries WITHOME lib wObj lib WITHOME lib wPyramid lib fsdStep Generate a new level in a FSD pyramid a h F w Description The fsdStep operator performs one iteration in the generation of a FSD pyramid see fsd The image output on the Gaussian port is equivalent to applying the reduce operator to the input image The image output on the Laplacian port is the difference between the input image and the filtered version of that image The Final output port is a flag which is set to 1 if this last iteration is the final step in a FSD pyramid generation The flag is 0 otherwise The final iteration occurs when the output image size is less than mini
65. Hough transform of the input image The direction parameter specifies whether the forward or reverse transform is performed The operator accepts unsigned 8 or 16 bit images as input and produces an unsigned 16 bit image as output The forward transform converts lines in a binary input image to points in output image It generates a greyscale image output where each output pixel value represents the number of pixels on the corresponding line in the input image In the output image the x axis represents the angle of the normal vector to the line in the input image with respect to the x axis positive angle values indicate clockwise rotation The angle range is specified by the startAngle and endAngle parameters in increments specified by the angleIncr parameter The angle parameters must be positive integers and the maximum angle range is 180 degrees The y direction represents the radial distance of the line from the center of the input image in image pixel units The height of the transform image is approximately sqrt W W H H where W is the width and H the height of the input image Radial distance zero is at one half the height of the Hough output image and positive radial distances extend upward negative y from that point The reverse transform converts points in the input image to lines in the output image It is not a true inverse in that when it is applied to a Hough tranformed image it does not reproduce the original image
66. Machine Intelligence July 1987 Border conditions are handled by assuming all pixels outside of the input image have zero value Note that when the center of the kernel is zero it is possible for dilation to clear an output pixel even if the input pixel is set Greyscale dilation is performed by computing the maximum value within every neighbourhood of the input image As with binary dilation only those pixels whose corresponding kernel bits are set are used to determine the local maxima Pixels beyond the image boundary are assumed to be the minimum value possible for the type of image being processed This means that they will have no effect when dealing with a maximum calculation Given a grayscale image J the following identity holds for binary dilate and gray scale dilate threshold gray scale dilate I binary dilate threshold 1 Parameter Information type Type int Default 0 Values Binary 0 Grayscale 1 kernelType Type int Default 0 Values Rectangular 0 Arbitrary 1 width Type int Default 7 height Type int Default 7 kernel Type Corlmage Default 1 1 1 1 1 1 1 1 1 optimization Type int Default 1 Values preciseness 0 speed 1 C Prototype CorOpRtn cor_dilate CorImage In CorImage Out int type int kernelType int width int height CorIntImage kernel int optimization CorOpRtn cor_dilate_consume CorImage In int type int kernelType int width in
67. Result CorString field Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib putField Insert a new field value into input object mii E ET ae Description The putField operator allows the user to insert a new value into a particular field in a complex object or vector of complex objects The operator s inputs are the object to be modified and the new field value Object types that can be modified include the built in types Point WitImage Vector and List extended types such as Real point Graphic and Feature and user defined types The field to be modified is specified by name in the field parameter using a syntax similar to the C language If the new field value is not the same type as the specified field in the complex object it will if possible be cast to that type The valid casts and rules for casting are described in the cast operator help When floating point values floats doubles or Real point are cast to integer types by the putField operator the values are rounded to the nearest integer When a value is cast to a type with a smaller range than the input type the value is truncated if it is outside the range If the input type cannot be cast to the field type an error occurs Fields in the WiT built in types that may be of use include x and y in Point W image width H image height bpp bytes per pixel and type in image and data size nco
68. SubImages input port A rectangle or point type graphic object or in the case of a vector of subimages a vector of graphic objects is input at the Locations input port If these inputs are vectors they must contain the same number of elements The Image input receives an image or vector of images The subimages are offset to the position specified by the corresponding graphic and are written over the image If a vector of images is received then all of the subimages are written over each image in the vector If the overlay parameter is set to Yes the original image values will be written to pixels not covered by any subimage In this case both the image and the subimage must be of the same type If the overlay parameter is set to No these pixels are set to zero Regions in the output image where subimages overlap will be set to the pixel values of the subimage that is later in the vector of subimages Parameter Information overlay Type int Default 1 Values off 0 on 1 C Prototype CorOpRtn cor_insert CorObj Image CorObj SubImages CorObj Locations CorObj Out int overlay Headers include WITHOME h wPixel h Libraries WITHOME lib wObj lib WITHOME lib wPixel lib putCol Put values into a column of pixels in an image at a LW Description The putCol operator sets pixel values in a specified column of the input image to the corresponding values in the input vector The column is given
69. The edgeDirection operator finds edges in images by detecting areas of high slope using Prewitt or Sobel gradient type operations Then the edges are shrinked in the perpendicular direction The operator produces two output images representing the magnitude and direction of the edges vectors at each pixel The magnitude output image is 8 bit unsigned integer image for magnitude and 8 bit signed image for orientation if degrees paramiter is 256 and float image if it is 360 The top output is the magnitude image and the bottom output is the orientation The direction is in the conventional units in the range 128 to 127 or in degrees range from 180 0 to 180 0 The output parameter can be used to select either a grayscale output image in which the output image represents a gradient magnitude or a binary output image All the edges below threshold are not included Parameter Information type Type int Default 0 Values Prewitt 0 Sobel 1 Roof 2 output Type int Default 0 Values Binary 0 Gray Level 1 threshold Type float Default 0 degrees Type int Default 0 Values 360 0 256 1 Demo iGraph Operators Filters conv2d igr C Prototype CorOpRtn cor_edgeDirection CorImage In CorImage magnitude CorImage orientation int type int output float threshold int degrees Headers include WITHOME h wFilter h Libraries WITHOME lib wObj lib WITHOME lib wFilter lib entropy Comput
70. The match values are compared to the value of the threshold and sufficient thresholdparameters which are set between 0 and 1 The numToMatch parameter applies a limit to the number of matches collected for output Regions of the target for which the match value exceeds the value of the sufficient threshold parameter are considered to be matches to the template If numToMatch matches that exceed this threshold are found the search stops and the matches are output Regions of the target for which the match value exceeds the value of the threshold parameter are considered to be potential matches to the template If the entire search region is searched without finding numToMatch matches over the sufficient threshold the best matches over the threshold value will be output The match value locations are output in decreasing order of match value Reduced sampling of the template can be specified using the parameters on the SubSampling sub panel The templateSampling parameter value specifies the pixel sampling period in the template used to calculate the match value at each pixel in the target The templateXOffset and templateYOffset parameters specify where in template sampling starts The targetSampling parameter value specifies the pixel sampling period in the target image The targetXOffset and targetYOffset parameters allow the samples to be taken at various points in the sample grid Parameters on the Result sub panel control the some
71. Type float Default 255 Demo iGraph Operators Transforms FourierFilter C Prototype CorOpRtn cor_equalize CorImage In CorImage Out float fromLo float fromHi float toLo float toHi Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib false Generate a zero valued integer object ae Description The false operator is used to convert any object received at its input port into a zero valued integer object This operator is useful for input to a variety of flow control operators requiring an integer value Demo iGraph Operators Calculate CollectAverage C Prototype CorOpRtn cor_false int Out Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib 19 fastAluOp Faster aluOp requires same size images Description The fastAluOp performs basic pixel by pixel arithmetic and logical operations on two input images Arbitrary expressions written in the C language syntax which may also include the pixel location can also be performed The input images may be grayscale or color images If both input images are grayscale images the output image type is selected according to the following criteria Grater nea Image Types m Image Type if either lif either input is float is float output is float is float K ee input is 16 output is 16 bit if either input is output is signed sign
72. a re calibration is necessary The proportion output from calibrateRatio can be used to scale all the old coordinates to fit in the new transformation C Prototype CorOpRtn cor_calibrateRatio CorCalibrateInfo calibrateInfoOld CorCalibrateInfo calibrateInfoNew float proportion Headers include WITHOME h wCalibration h 32 Libraries WITHOME lib wObj lib WITHOME lib wCalibration lib calibrateScale Scale calibration transformation Description The calibrateScale operator changes a calibration transformation to scale the object multiplying all coordenates by scale Parameter Information scale Type float C Prototype CorOpRtn cor_calibrateScale CorCalibrateInfo calibrateInfo CorCalibrateInfo calibrateInfoScaled float scale Headers include WITHOME h wCalibration h Libraries WITHOME lib wObj lib WITHOME lib wCalibration lib moveOrigin Apply shift and rotation to calibration transformation atl s Description The moveOrigin operator changes a calibration transformation to translate the origin dx dy to the specified location and to rotate angle the coordinate system about the origin Parameter Information dx Type float dy Type float angle Type float C Prototype CorOpRtn cor_moveOrigin CorCalibrateInfo calibrateInfo CorCalibrateInfo xtransformedCalibrateInfo float dx float dy float angle Headers include WITHO
73. all non zero pixels are considered to be equivalent and the run lengths will represent runs of non zero pixels When it is set to labeled pixels with the same value are considered equivalent and run lengths represent runs of equal valued non zero pixels maxBlobHeight The maxBlobHeight parameter sets an upper limit on the height of blobs produced by the operator When the information derived using the webGetBlobFeatures operator is used with images produced by the webWaterfall operator the maxBlobHeight value should be less than the height parameter value of the webWaterfall operator by at least the height of the images in the input image stream to ensure that the blobs are contained within a single waterfall image reset The reset parameter can be used to release all active blobs and reset the running line count to zero at the end of 275 the current image When a reset function is required this parameter is usually used as an input parameter When a zero is received the blobs are built up as usual When a one is received all blobs in the current image are output even if they reach the bottom of the image and after processing for the current image is complete the starting line number is reset to zero The webLineCounter operator can be used to provide the required control signals When no reset function is required this parameter is set to No In this mode the line count will continually increase as images are received unt
74. allows the specification of the type of object to be created and its value Parameter Information value Type CorObj C Prototype CorOpRtn cor_crObject CorObj Out CorObj value Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib crVector Create a vector of same type as input Description Create a vector of the same type as the input object For example if the input object is an integer then the vector created will be a vector of integers The number of elements is specified by the num parameter All the elements are set to zero Parameter Information num Type int C Prototype CorOpRtn cor_crVector CorObj sample CorVector vec int num Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib createRect Create a rectangle graphic object 159 ij je Description The createRect operator creates a rectangular Graphic object from the coordinates specified by the p1 and p2 parameters The two points p1 and p2 provide the coordinates for two vertices of a rectangle All coordinates must be positive integers This operator may be used with the extract operator to extracting an exact area of interest from an image for example Parameter Information pl Type CorPoint Default 0 0 p2 Type CorPoint Default 64 64 C Prototype CorOpRtn cor_createRect CorGraphic Rect CorP
75. also accept the obsolete OBJ_ object name format Multiple user object types can be specified by separating their names by spaces Parameter Information type Type int Default 0 Values bitwise OR byte 1 ubyte 2 short 4 ushort 8 int 16 uint 32 hex 64 float 128 double 256 String 512 Point 1024 image 2048 vector 4096 linked list obsolete 8192 Graphic 16384 Geometric 32768 user 65536 char 131072 hex8 262144 hex16 524288 File 1048576 user Type CorString Default OBJ_GENERIC image Type int Default 0 Values bitwise OR unsigned 8 bit 1 signed 8 bit 2 unsigned 16 bit 4 signed 16 bit 8 float 16 complex 32 RGB 64 HSV 128 Yuv 256 graphic Type int Default 0 Values bitwise OR point 1 line 2 polyline 4 polygon 8 rectangle 16 circle 32 text 64 arc 128 C Prototype CorOpRtn cor_filter CorObj In CorObj Equal CorObj NotEqual int type CorString user int image int graphic Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib flattenVector Convert vector of vectors to vector Pakeha afiiHef F Description The flattenVector operator converts a 1 D object vector in which each of the elements is a 1 D object into a flat 1 D If the setNCols parameter is set to No the output is a single 1 D object formed by concatenating the elements of the input 1 D object If
76. and a control signal that indicates an image represents the last image the current large image space Inputs In The operator takes an image or an integer value as input Outputs BlockSize The height of the input image or the value of the input integer is output at the block BlockSize output YOffset The line number of the first line of the image relative to the first image of the current large image space is produced at the YOffset output Reset A one is produced on the Reset output when the last image each group of input images of the size specified in the blocksPerImage parameter is received A zero is produced when any other image is received Parameters blocksPerImage The blocksPerImage parameter specifies the number of images that make up the larger image space After receiving this number of images the YOffset line counter is reset The operator also produces a reset signal 1 at the reset output Parameter Information blocksPerlmage Type int Default 32 C Prototype CorOpRtn cor_webLineCounter CorObj In int BlockSize int YOffset int Reset int blocksPerImage Headers include WITHOME h wWeb h Libraries WITHOME lib wObj lib WITHOME lib wWeb lib webPlotFeatures Convert feature vectors to graphic descriptions in waterfall image Description The webPlotFeatures operator converts the input feature vector element or feature vector generated from blobs in one frame of
77. and y coordinates of the location in the matching region in the target image corresponding to the center pixel in the pattern image For even dimension width or height pattern images the center pixel location is rounded up to the next integer value Only the best match at each 260 location is reported The matches are sorted in order of descending score Scores The match scores corresponding to the locations in the Locations output vector are produced as a vector of float values Indexes The Indexes output is a vector of integers in which each element is the index number of the pattern in the PatModel that produced the match at the corresponding location in the Locations output vector Parameters scale The scale parameter is used to specify resolution reduction factor for the search It can be set to powers of two from one to 32 range The range parameter specifieds the size of the neighbourhood searched around each input point The search will extend at most the specified number of pixels in the x and y directions It can be set to powers of two from one to 32 rejectThresh The rejectThresh parameter is used to set a rejection threshold for match scores below which no matches will be considered The parameter must be set to a value from zero to one where zero indicates no correlation between the pattern and target and one indicates a perfect match If no part of the target image matches the pattern sufficiently well emp
78. aspects of the output format The result parameter is used to select the graphic object that locates the matched subimage regions in the target image Selecting center or upper left will produce a point located at the center or upper left of each region while selecting bounding box will produce a rectangle corresponding each regions bounding box The vector of graphic objects can be sent to the overlay operator for display on the input image or used to extract a subimage from the target image The result parameter also controls how the template is overlayed relative to the selected target pixels as described above The vicinity parameter is used to compare matched objects with similar position If a new match is found which is within vicinity pixels in x and y of a previously found match point then only the better of the match points is recorded The red green and blue parameters can be used to control the color of the output graphics The default color is red Parameter Information method Type int Default 1 Values Sum of Diff 0 Norm Corr 1 Binary 2 correlation Type int Default 0 Values positive 0 negative 1 both 2 threshold Type float Default 0 5 sufficient Type float threshold Default 1 numToMatch Type int Default 5 templateSampling Type int Default 1 templateXOffset Type int Default 0 templateYOffset Type int Default 0 targetSampling Type int Default 1 targetX Offset Type int Defau
79. bjects for which the first field values are equal or within a tolerance range are sorted on the second field value A typical application of this operator is to sort Feature vectors derived from images of lines of text so that the lines are sorted in y and the characters in each line are sorted in x The first field to be sorted on is specified using the field1 parameter The tolerance parameter can be used to specify a range over which the first field values will be considered equal The second field value is specified using the field1 parameter The fields must be one of the numeric or character basic types The vector can be sorted in ascending or descending order for each field as specified by the direction1 and direction2 parameters The output parameter is normally set to object which produces an output vector containing the the objects from the input vector in sorted order The output parameter may be set to index in which case the output is an integer vector where each element is an index into the input vector for the object rather than the object itself It should be noted that in general using the first field value with a non zero tolerance range does not result in an ordering of the objects That is for three objects a b and c it is possible that a lt b b lt c and c lt a Heuristics are used that provide the expected result in cases where the first field values are clustered with some dependence on the seco
80. by the col parameter Parameter Information column Type int Default 0 C Prototype CorOpRtn cor_putCol CorImage In CorVector ColumnData CorImage Out int column Headers include WITHOME h wPixel h Libraries WITHOME lib wObj lib WITHOME lib wPixel lib putPixel Put a pixel value into an image Ea Description The putPixel operator sets a single pixel in an image to a specified value The pixel that is set to the new value is located by the x y coordinate given in the xy parameter Parameter Information xy Type CorPoint Default 0 0 C Prototype CorOpRtn cor_putPixel CorImage In CorObj Pixel CorImage Out CorPoint xy Headers include WITHOME h wPixel h Libraries WITHOME lib wObj lib WITHOME lib wPixel lib putRow Put values into a row of pixels in an image 179 Description The putRow operator sets pixel values in a specified row of the input image to the corresponding values in the input vector The row is given by the row parameter Parameter Information row Type int Default 0 C Prototype CorOpRtn cor_putRow CorImage In CorVector RowData CorImage Out int row Headers include WITHOME h wPixel h Libraries WITHOME lib wObj lib WITHOME lib wPixel lib putVector Put values into pixels along an arbitrary line segment Description The putVector operator inserts the values in the input
81. character of the string The user can hit the cancel button to cancel the operator In this case no output is produced The label parameter is used as a label on the dialog Parameter Information label Type CorString Default Hit a key Placement Type int Default 0 Values bitwise OR Fix position 1 X Type int Default 0 Y Type int Default 0 C Prototype CorOpRtn cor_iKkKey CorString Out CorString label int Placement int X int Y Headers include WITHOME h wDisplay h Libraries WITHOME lib wObj lib WITHOME lib wDisplay lib iMeas Interactive image measurement 60 Description The iMeas operator allows image measurements such as the length of a path or the area and average pixel value in a region of interest to be made interactively The image region to be measured is specified by a graphical object drawn on the image using the mouse and the Graphics Editor as described for the getData operator an a ity od Lege TH J e Pom ome Lee ara E imme Er aE hatri TA Measured values related to the graphic object and to the pixel values in the graphic are displayed in a popup panel The measured values related to the graphic include the length of the graphic perimeter in the case of solid geometric objects such as polygons or circles the area of the graphic and the bounding box height and width of the graphic These values are scaled by th
82. characters represented by the RSS14 If no valid RSS14 is found an empty string is outputed If the ROI is a vector of rectangles then the operator seeks for a RSS 14 in each ROI The output is a vector of String objects If a valid RSS14 is found in a ROI the corresponding element of the output vector contains the decoded string If no valid RSS14 is found in a ROI the corresponding output vector element is the empty string Parameters noise The noise parameter specifies the minimum low to high or high to low change in pixel value necessary to constitute the edge of bar type The type parameter specifies the RSS14 barcode type If parameter value is regular then RSS 14 belongs to the RSS 14 RSS 14 Truncated RSS 14 Stacked or RSS 14 Stacked Omnidirectional otherwise it belongs to RSS 14 Limited Parameter Information noise Type int Default 5 type Type int Default 0 Values regular 0 limited 1 Demo iGraph Operators Matrix POSTNET igr C Prototype CorOpRtn cor_getRSS14 CorImage Target CorObj ROI CorObj Decoded int noise int type Headers include WITHOME h wMatrix h Libraries WITHOME lib wObj lib WITHOME lib wMatrix lib getUPCa Finds and decodes UPCa bar code E sN Description The getUPCa operator accepts a binary or grayscale image and one or more rectangular regions of interest ROIs then searches the ROIs for UPC version A barcodes If a barcode is found in a r
83. effect Controls the visibility of pixels in the volume based on pixel value Opacity changes require that the entire volume be reclassified so these changes may be slow for large volumes depending on available memory and processor speed If Threshold is selected values below the threshold are transparent values above the threshold are opaque The threshold level is set using the lower level arrow on the Opacity ramp control If Ramp is selected pixel values between the upper and lower level arrows of the Opacity ramp control are semi transparent with opacity varying linearly from transparent at the lower level to opaque at the upper level Values below the lower level arrow are transparent and values above the upper level arrow are opaque If Custom is selected you can define an arbitrary mapping of pixel values to opacity using the Map Editor The opacity changes are not applied to the Controls the viewpoint from which the volume is observed Refer to the chapter Three Dimension Viewing Widgets in the Reference manual for details Viewing Allows viewing the interior of a volume The slider specifies the position of a plane perpendicular to one of the volume axes as a percent of the volume size in that direction All pixels in front of the cut plane are transparent The cut planes are limited to planes perpendicular to the current primary viewing axis which may chang
84. error occurs if there is no element n in the vector Field names of extended or user defined types can be seen by viewing an object of the type using a probe or the display operator For example displaying a Feature object produces a list of the feature s fields id xmin xmax and their values The names as displayed including case can be used to access these fields Substructures embedded within structures can be specified using either or gt notation Fields can be nested to an arbitrary depth For example the Graphic object has a field called geom which is itself a Geom object that includes the field plist This field can be accessed by using geom plist or geom gt plist as the field parameter The plist field is the vector of points that describe the graphic object The number of elements in the plist vector can be extracted by specifying geom plist size The coordinate values of a given point in the graphic object can be accessed using the vector notation described above For example geom plist data 0 x specifies the x coordinate of the first point in the vector The getField operator can also extract fields from the elements of a vector of complex objects and collect these fields into a vector of the field type for output To specify this mode the field parameter is prefixed with a For example an integer vector of the size in pixels of objects in a binary image can be extracted directly from the vector of
85. executed in two modes The user may choose the mode by using the optimization parameter If the parameter value is speed then execution goes way faster otherwise operator delivers more precise results Binary or gray scale opening is selected using the type parameter The input for binary opening must be an 8 bit unsigned image The result is a binary 8 bit unsigned image For gray scale opening the input may be any integer or float gray scale image or color in the case optimisation is speed The output image is a gray scale image of the same type as the input image The kernelType parameter is used to select the type of structuring element or neighbourhood specifying kernel used The structuring element may be specified as a rectangular region with the size of the region is given by the width and height parameters or it may be specified as an arbitrary region in which case the neighbourhood is 149 specified using the kernel parameter as described in the dilate or erode operator documentation Opening by a rectangular structuring element is usually faster than opening by an arbitrary element particularly for larger neighbourhoods One common exception is opening by 3x3 structuring elements where the arbitrary kernel usually is faster Parameter Information type Type int Default 0 Values Binary 0 Grayscale 1 kernelType Type int Default 0 Values Rectangular 0 Arbitrary 1 width Type int Default 7 height
86. fraction of the range from the image minimum value to the image maximum value so that a value of O gives a threshold equal to the image minimum and a value of 100 gives a threshold equal to the image maximum For minima the value is set to the 94 fraction of the range from the maximum to the minimum so that a value of 0 gives a threshold equal to the image maximum and a value of 00 gives a threshold equal to the image minimum If the value calculated using relThreshold is greater than the threshold for maxima or less than the threshold value for minima then it replaces the threshold value The maxPeaks parameter specifies the maximum number of peaks to be output If more peaks than the maximum number of peaks are found the peaks with the highest pixel values are output for maxima and the peaks with the lowest pixel values are output for minima If the subpixel parameter is set to Yes pixel values in the neighbourhood of the maximum or minimum value pixel will be used to estimate a sub pixel location for the peak If subpixel is set to No and the peak is includes more than one pixel the location will be the upper left pixel for maxima and the lower right pixel for minima Parameter Information threshold Type float relThreshold Type float maxPeaks Type int Default 10 type Type int Default 0 Values maxima 0 minima 1 subpixel Type int Default 0 Values No 0 Yes 1 Demo iGraph Operators Measurement Cir
87. grabber The port parameter selects which port to configure A port is a collection of 1 or more I O lines commonly 8 or 16 that are read or written in a single operation Normally a particular port has read only or write only lines For example port 0 may be read only while port 1 may be write only The number of available ports the number of lines per port and the direction of the lines are all hardware dependent characteristics See the documentation for your frame grabber for details The direction parameter controls whether the lines chosen by the bits parameter of the selected port are to be configured for input reading or output writing The bits parameter controls which lines of the selected port are to be configured The value of bits is interpreted as a bit field the lines corresponding to the on bits are the ones to be configured For example if the value of bits is 170 bit pattern 10101010 then every other line is to be configured for reading or writing depending on the value of the direction parameter If the hardware supports ports which can have lines for both reading and writing the acqloSetup operator can be executed twice once to configure the read lines and once for the write lines Normally all lines in a port are read only or write only lines In this case the bits parameter should be set to the value corresponding to a bit pattern of all 1 s For example a port with 8 lines would use a v
88. h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib getRoi Extract ROI from an image 163 Description The getRoi operator extracts the image ROI from the input image and outputs the ROI as a rectangle Graphic The color of the output rectangle can be set using the red green and blue parameters Parameter Information red Type int Default 255 green Type int Default 0 blue Type int Default 0 C Prototype CorOpRtn cor_getRoi CorImage i0 CorGraphic ROI int red int green int blue Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib getTiles Convert an image to a vector of sub images Description The getTiles operator decomposes an image into a vector of non overlapping sub images The operator accepts any image type and produces a vector of images of the same type 164 The set parameter value determines the grid used to divide the image into sub images It may be set to the divide the image by the size of the resulting sub images or by the number of rows and columns into which the image is divided When the parameter is set to size parameters are enabled to specify the width and height of the sub images When the parameter is set to number parameters are enabled to specify the number of columns nCols and number of rows nRows into which the image is divided The remainder parameter determines what happens wh
89. h Libraries WITHOME lib wObj lib WITHOME lib wSegment lib localPeaks Find local maxima or minima in images Description The localPeaks operator finds local maxima or minima in an image and produces a binary output image in which the non zero pixels represent the maxima or minima found The peakType parameter is used to specify either minima or maxima A local maximum may be a single pixel or a plateau A single pixel maximum has a value greater than all adjacent pixels A plateau is a contiguous group of pixels with equal value where all pixels not in the group and adjacent to any pixel in the group have a value less than pixels in the group The definition of local minima is similar with values less than those of adjacent pixels The connectivity parameter is used to specify which pixels are considered adjacent to another pixel If it is set to 4 the four edge connected 194 neighbours are considered the adjacent pixels If it is set to 8 the four diagonal neighbours are also included Parameter Information peakType Type int Default 1 Values minima 0 maxima 1 connectivity Type int Default 1 Values 4 0 8 1 Demo iGraph Operators Segmentation LocalPeaks C Prototype CorOpRtn cor_localPeaks CorImage In CorImage Out int peakType int connectivity Headers include WITHOME h wSegment h Libraries WITHOME lib wObj lib WITHOME lib wSegment lib localThresh L
90. high transition occurs where a pixel value is greater than or equal to zero and the value of the preceding pixel in the row is less than zero subject to the contrast test described below Similarly a high to low transition occurs where a pixel value is less than or equal to zero and the value of the preceding pixel in the row is greater than zero When all zero crossings are desired both low to high and high to low transitions are marked A contrast test ensures that the amplitudes of the high and low pixel values defining the zero crossing are more than the value specified by the contrast parameter away from zero The contrast test searches backward and forward from a potential zero crossing point to ensure the given contrast is present within the zxwidth For example if lo gt hi transitions are desired then for each such transition found the algorithm will search backwards along the negative lobe for not more than zxwidth pixels for a pixel value more than the given contrast below zero If this condition is met then the algorithm searches forward along the positive lobe for not more than zxwidth pixels for a pixel greater than the given contrast If this condition is met then the zero crossing is marked otherwise it is rejected When a zero crossing is marked the search for zero crossings resumes from the pixel at which the forward search criterion was met This means that zero crossings within the zxwidth range may be ignored
91. image outputs when 4 camera switching is enabled For 2 camera switching only the top 2 outputs are used For 3 camera switching only the top 3 outputs are used C Prototype CorOpRtn acqSwitchGrab CorImage Cameral CorImage Camera2 CorImage Camera3 CorImage Camera4 Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqWriteLut Write an input LUT on frame grabber Description Write an input LUT on frame grabber Parameter Information filename Type CorFile Default sampleLut C Prototype CorOpRtn acqWriteLut CorIntVector Lut CorFile filename Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqZoom Set hardware scale on frame grabber a Description The acqZoom operator is used to scale an image in hardware as it is being grabbed using the acquire operator The amount of scaling is determined by the x and y parameters A positive integer is used to magnify the image whereas a negative integer is used to reduce the image For example setting the parameters to 2 will cause the image to be reduced to half its size whereas a setting of 2 will cause the image to be magnified by double its size Parameter Information x Type int Default 1 y Type int Default 1 C Prototype CorOpRtn acqZoom int x int y Headers include WITHOME h wVideo h Librari
92. image size is specified by the width and height parameters The image type can be selected from among unsigned 8 or 16 bit integer image float image or RGB color image using the type parameter The image values are set using the fillType parameter Selecting the fillval option will set all image pixels to the value specified in the fillval parameter The value will be clipped if it is outside the range of the selected image type Selecting the h ramp option will generate a ramp in the x direction The pixel values increase linearly from 0 in the leftmost column of the image to a maximum at the rightmost column The maximum value is 256 for 8 bit integer float and RGB images and 65536 in 16 bit integer A similar ramp is generated in the y direction when v ramp is selected Parameter Information width Type int Default 100 height Type int Default 100 type Type int Default 0 Values 8 bit 0 16 bit 1 float 2 color 3 fillType Type int Default 0 Values fillval 0 h ramp 1 v ramp 2 fillval Type int Default 0 C Prototype CorOpRtn cor_crimage CorImage Out int width int height int type int fillType int fillval Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib crObject Create an object of any type a a Description The crObject operator creates an object of any type including user created objects The value parameter
93. images can only be offset downward from the top of the image The offset spaces are filled with the lines from the bottom of the corresponding portion of previous images These spaces are filled with zero valued pixels in the first image in the sequence Offset The Offset output is an integer value that gives the line number of the first row of the output image relative to the first row in the first camera image in the sequence of images Parameters offset reset validRange The validRange parameter is used to specify the start and end pixel for the valid non overlapping horizontal range for each camera The number of elements in this Point vector implicitly specifies the number of cameras The start and end values must be within the portion of the image captured by the corresponding camera The offset parameter is a vector of integers in which each element represents vertical offset of the corresponding region in the input image The vector must contain the same number of entries as the validRange parameter vector clear The reset parameter is used to reset the line count to zero When no reset function is required the parameter should be set to No When the reset function is required the parameter is used as an input parameter When a zero is received the line count is incremented as usual When a one is received the line count is reset to reference the first row of the next image received after processing the curre
94. images to produce two output images of the same width and height The offset parameter is used to specify the offset applied to image B relative to image A The dimensions of the output images are determined from the input images as specified by the sizeOp parameter clip B B is clipped to fit in A expand A A is expanded to include B clip A and B A and B are clipped to their intersection center B on A B is centered on A and then clipped to A Clipping and expanding is performed after offsetting image B Values in the output images Aout and Bout are set to the values of the corresponding pixels in A and B respectively where they are defined and to zero elsewhere The outputType parameter allows the user to specify the type of WiT image produced The default option causes a default output image type to be selected according to the following criteria Input Image Types Output Image Type if either input is color output is color else if either input is complex output is complex else if either input is float output is float else if either input is 16 bit output is 16 bit if either input is signed output is signed Color images are converted to other types by calculating a gray scale luminance image equivalent to the Y channel of the Yuv color system which is cast to the required type Gray scale images are converted to color images by converting them to gray R G
95. in the font The vertices of a polygon represent the sample points of the character perimeter The label field of a Graphic contains the label associated with the character in the font model The serial field contains the character s index number which may be required by the ocrModify operator Parameters width and height The width and height parameters specify the minimum width and height of the space in which each reconstructed character graphic is centered The width of the space is set to the greater of the maximum character width in the font model and the value of the width parameter and the height is set to the 245 greater of the maximum character height and the value of the height parameter offset The positioning of the polygons is controlled by the offset parameter When offset is set to None each polygon is positioned so that it is centred on an image the size of the reconstructed character space as described above When the offset parameter is set to Tile the character polygons are positioned in rows suitable for overlay on a single image Each character is centered on a tile the size of the reconstructed image space When the offset parameter is set to Original the character polygons are placed in the position of the character in the original binary image from which the perimeter was derived cols The number of polygons in each row in the tiled offset mode is set by the cols parameter If the cols parameter
96. include WITHOME h wPyramid h Libraries WITHOME lib wObj lib WITHOME lib wPyramid lib Segmentation The Segmentation library produces binary or labeled images from grayscale source images The library includes thresholding operators and other operators that divide an image into regions based on pixel values It also includes operators that mark distinctive pixels in the image adaptThresh Threshold an image relative to image mean pixel value Description The adaptThresh operator applies thresholding to a grayscale image using thresholds based on pixel values in the input image For grayscale images the thresholds are applied directly to the values of image pixels For color images the thresholds are applied to a luminence or intensity value for each pixel the Y channel for Yuv images the V channel for HSV images and a luminence value derived from all three channels equivalent to the Y value for RGB images The operator function is same as that of the thresh operator except that values corresponding to the lo and hi parameters of the thresh operator are calculated from the image pixel values Each threshold value is set to the mean image pixel value plus an offset The method of calculating the offsets is specified by the offsetMethod parameter The low and high offset values are determined using the loOffset and hiOffset parameters respectively 189 When the offsetMethod parameter is set to absolute th
97. input and produces an output image of the same type as the input Parameter Information kernel Type Corlmage Default 0 0 1 1 1 0 0 0 1 1 1 0 1 0 1 1 1 1 0 0 0 1 0 1 1 1 1 1 0 gt 1 1 1 1 0 0 gt ato ty ty ty te Me MMe threshold Type float Default 10 overlay Type int Default 0 Values off 0 on 1 Demo iGraph Operators Morphology skeleton igr C Prototype CorOpRtn cor_tophat CorImage In CorImage Out CoriIntImage kernel float threshold int overlay CorOpRtn cor_tophat_consume CorImage In CoriIntImage kernel float threshold int overlay Headers include WITHOME h wMorpho h 154 Libraries WITHOME lib wObj lib WITHOME lib wMorpho lib Objects The Objects library provides functions to manipulate WiT objects and handle elements in vectors and fields in compound objects The library includes operators that create objects operators that read and write to fields in compound objects and operators that insert extract or reorder elements of vectors addElem Add an element to a vector Description The addElem operator appends the object at the NewElement input port to the vector at the Vector input port The type of the object to be added must be the same as the type of objects currently stored in the input vector The position at which the element is added can be specified to be the first the last or an intermediate eleme
98. integer amounts in x and y using bicubic interpolation Inputs In The input must be a gray scale integer or float image Color and complex images are not supported by this operator Outputs Out The output image is the same size and type as the input image Pixels in rows and columns outside the shifted input image are set to zero Parameters dx and dy The dx and dy parameters specify the amount of shift in the x and y directions respectively A positive value of dx shifts the input image to the left and a positive value of dy shifts the input image up method The method parameter allows the user to choose one of two interpolation methods If the parameter is set to Derivatives the operator uses the values and derivatives at the four pixels surrounding each interpolation point to determine the value at the interpolation point The derivatives are calculated from pixel value differences If the parameter is set to Lagrange the operator fits third order polynomials to pixel values in the neighbourhood of the interpolation point to obtain a value Parameter Information dx Type float Default 0 5 dy Type float Default 0 5 method Type int Default 0 Values Derivatives 0 Lagrange 1 BiLinear 2 C Prototype CorOpRtn cor_patInterpolate CorImage In CorImage Out float dx float dy int method Headers include WITHOME h wSearch h Libraries WITHOME lib wObj lib WITHOME lib wSearch lib pat
99. inverted using the fomula nz z The output image is the same type as the input image Demo iGraph Operators Calculate HistEqualize C Prototype CorOpRtn cor_invert CorImage In CorImage Out CorOpRtn cor_invert_consume CorImage In Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib makeLut Generate a lookup table aca Description The makeLut operator produces a lookup table LUT by parsing a C expression representing the desired mapping function The LUT is used as an input to the applyLut operator to implement point operations on images The LUT provides a mapping from the integers A over the range of the pixel values in the expected input image type for the applyLut operator to pixel values in the output image The mapping is specified by a function f A as defined by the parameter expression The userConstant parameter allows the user to specify a constant for use within the expression For more on the syntax and usage of expression see calc The LUT is output from the operator as a vector of values in which the size of the vector and type of the values is determined by the inputType and outputType parameters The inputType parameter must be set to the type of input images expected to be applied to the applyLut operator It may be set to one of the integer image types ubyte byte ushort or short The size of the output vector is
100. is ceil low floor high so if the low and high values are integers they are both included in the range The range for floats is low high that is the low value is included in the range the high value is not The random number sequence used to generate the numbers can be controlled using the useSeed seed and reset parameters A random number seed is generated the first time an operator executes or when the reset parameter is set to Yes starting a new random number sequence When the reset parameter is set to No the random number sequence is a continuation of the sequence used the last time the operator executed When the useSeed parameter is set to No the initial random number seed is generated from a value derived from the system clock so a different random number sequence is produced each time the graph is run When the useSeed parameter is set to Yes the value specified in the seed parameter is used to initialize the random number sequence producing repeatable sequences of random numbers The seed value is a three element vector of short integers The operator s NextSeed output value is the seed value that will continue the random sequence used by the operator Parameter Information useSeed Type int Default 0 Values No 0 Yes 1 seed Type CorVector Default 1 0 0 number Type int Default 10 type Type int Default 0 Values int 0 float 1 low Type float Default 0 high Type float De
101. is usually used as an input parameter When a zero is received the blobs are built up as usual When a one is received all blobs in the current image are output even if they reach the bottom of the image and after processing for the current image is complete the starting line number is reset to zero The webLineCounter operator can be used to provide the required control signals When no reset function is required this parameter is set to No In this mode the line count will continually increase as images are received until the graph stops Blobs which reach the bottom of the last image processed will not be output by the operator resetBlobId The blob id number can be set to zero with the resetBlobId parameter This will set the next blob id to zero after processing the current image This parameter is usually set to No or used as an input parameter It may be tied to the reset parameter when both are input parameters in which case a new set of id numbers will be started every time the operator is reset Parameter Information connected Type int Default 1 Values 4 0 8 1 type Type int Default 0 Values binary 0 labeled 1 maxBlobHeight Type int Default 0 reset Type int Default 0 Values No 0 Yes 1 resetBlobld Type int Default 0 Values No 0 Yes 1 C Prototype CorOpRtn cor_webGetBlobsRoi CorImage BinaryImage CorObj ROI CorBlobVector Blobs int connected int type int maxBlo
102. less than B A lt B A less than or equal to B A B A equal to B A gt B A greater than or equal to B A gt B A greater than B and A B A not equal to B The output image is a uint8 binary image Non zero values in the output image indicate pixels in which the values in the input images meet the specified condition The output image is the same size as the A input image The output image ROI is set to the intersection of the ROIs of the two input images Pixels values are compared only within the output image ROI An error tolerance can be specified when float images are compared for equality or inequality using the floatEqError parameter Floating point pixel values will be considered to be equal if the pixel value from image B is greater than or equal to the pixel value from image A minus the error tolerance and less than or equal to the pixel value from image A plus the error tolerance Parameter Information condition Type int Default 0 Values A lt B 0 A lt B B 2 A gt B 3 A gt B 4 A 1A B 5 floatEqError Type float Default 0 Demo iGraph Operators Calculate CollectAverage C Prototype CorOpRtn cor_compareImages CorImage A CorImage B CorImage Out int condition float floatEqError Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib comp xMult Multiply a complex image by a real or complex image z iT
103. lib 175 getCol Get a column of pixels from an image Description The getCol operator retrieves the vector of pixel data corresponding to the column of pixels in the input image The column is given by the col parameter Parameter Information col Type int Default 0 C Prototype CorOpRtn cor_getCol CorImage In CorVector Colvec int col Headers include WITHOME h wPixel h Libraries WITHOME lib wObj lib WITHOME lib wPixel lib getPath Get pixel values and coordinates of points on a path Description The getPath operator produces the floating point location of rasterized or equally spaced sample points along a path specified by a graphical object 176 and determines the image values at the sample points The operator has two outputs one corresponding to the path and the other to the image values on the path The upper output is a vector of Fpoint objects that give the location of the sample points in pixel coordinates as described below The lower output is a vector of floating point values at the sample point locations in the image input at the upper input port The path is specified by a graphical object that is received on the lower input port Any graphical object type other than text can be used The path associated with rectangles circles and polygons is the perimeter of the object Point objects return a single sample point The operation parameter specifies whether the p
104. magnitude of the gradient can be obtained using the edgeDetect operator Parameter Information type Type int Default 0 Values Prewitt 0 Sobel 1 degrees Type int Default 0 Values 360 0 256 1 optimization Type int Default 1 Values preciseness 0 speed 1 Demo iGraph Operators Filters grad igr C Prototype CorOpRtn cor_grad CorImage In CorImage Magnitude CorImage Orientation int type int degrees int optimization Headers include WITHOME h wFilter h Libraries WITHOME lib wObj lib WITHOME lib wFilter lib hipass1d One dimensional high frequency enhancing filter Description The hipassld operator applies one dimensional high pass filtering or sharpening in the x direction to the input image The effect of the operator is to enhance local contrast along each row in the image and to suppress slow changing values in each row One dimensional high pass filtering is faster than two dimensional high pass filtering Also since the filtering is only performed along the x axis original image information along the x axis is preserved First a smoothed image is calculated over the image as described for the lopassld operator The smoothed image is then subtracted from the original image to give the high pass filtered image The size of the kernel used for the smoothing is specified by the width parameter Increasing the kernel size lowers the cut off frequency of the filter and
105. mode parameter It may be set to max min or nearest When this parameter is set to nearest the value parameter is used to specify the value to which the element or field value is compared The output parameter is normally set to object in which case the Element output is the appropriate element from the input vector The remaining vector elements are produced on the Rest output The output parameter may also be set to index in which case the output is the integer index into the input vector for the object rather than the object itself This mode is most often used when selecting from vectors of simple numeric objects Parameter Information field Type CorString Default x mode Type int Default 0 Values max 0 min 1 nearest 2 value Type float Default 0 output Type int Default 0 Values object 0 index 1 C Prototype CorOpRtn cor_selectElem CorVector In CorObj Element CorVector Rest CorString field int mode float value int output Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib setRoi Set image ROI F Description The setRoi operator sets the ROI of an image to the rectangle specified by the ROI parameter The ROI parameter may be may a vector of two points or real points representing the upper left and lower right corners of the rectangle or it may be a rectangle graphic or geometry object or it may be an
106. must be selected The output image will be the same type as the color input image The aluOp will handle images with ROIs of different sizes The sizeOp parameter and the offset parameter control the processing ROI The offset parameter is used to specify an offset for image B relative to image A If the sizeOp parameter is set to clip B then the output image ROI will be set to the intersection of the ROI of A and the offset ROI of B If the sizeOp parameter is set to expand A then the output image ROI will be set to the bounding rectangle ROI of A and the offset ROI of B In this case input pixels outside the ROIs of the input images are given a value of zero so using addition as an example pixels in the intersection of the two ROIs will have the value of A B pixels in the ROI of A only will have the value of A pixels in the ROI of B only will have the value of B and pixels not in either input ROI will have a value of zero The output image is normally the same size as the A input image If necessary the output image will be made large enough to include the bounding ROI when sizeOp parameter is set to expand A Supported operations are A B A B A B A B 12 OR bitwise OR of A and B AND bitwise AND of A and B XOR bitwise XOR of A and B src use B where B 0 else use A overlay use B if A amp overlayMask and B 0 else use A expr utilize expression to determine result abs A B absolute value of differe
107. next non white space character in the input string If the characters do not match sscanf terminates These characters are read but not stored e Format specifications introduced by the percent sign Format specifications cause sscanf to read and convert the characters of the input string into a value of a specified type This value becomes one of the elements in the output vector A format specification can be suppressed by using the character following the character in which case the input field is read and not stored The following format specifications are supported by WiT d i decimal integers u unsigned integer o octal int returns the decimal value x X hexadecimal int returns the decimal value c C a string one character long s S string e E f g G float reads and stores until the first character that does not appear in the brackets e A reads and stores until the first character than does appear in the brackets e 6 read as a literal no assignment is made Sscan produces two outputs Out which holds the vector of elements converted and Count which holds the number of elements converted If there are more arguments in the input string than required by the format parameter the extras are discarded If there are not enough format parameters the extras are ignored Notes e Count is an optional output value always computed but reported only if it is conn
108. numToMatch int templateSampling int templatexOffset int templateYyOffset int targetSampling int targetXOffset int targetYOffset int searchRange int result int vicinity int red int green int blue Headers include WITHOME h wFind h Libraries WITHOME lib wObj lib WITHOME lib wFind lib xCorr Image cross correlation Description The xcorr operator performs the cross correlation of image A and image B The cross correlation operation is performed by taking the inverse discrete Fourier transform DFT of the product of the DFT of image A and the complex conjugate of the DFT of image B The operator accepts two integer or float images The width and height of image A must be greater than or equal to the width and height repectively of image B The output is a float image the size of image A where the pixel values are the cross correlation values for corresponding offsets of image B Values are only calculated at pixels where the entire image B lies within the boundaries of image A Demo iGraph Operators Find X Corr C Prototype CorOpRtn cor_xCorr CorImage A CorImage B CorImage Out Headers include WITHOME h wFind h Libraries WITHOME lib wObj lib WITHOME lib wFind lib 99 Fit The Fit library supports the production of higher level or derived geometric representations from image or geometric data The library includes operators that fit image or geometri
109. of ROIs The locations have a one to one correspondence with the label strings in the Labels output Locations The operator also outputs the locations of the characters found in the target image as either a Real point vector or a vector of Real point vectors depending on whether the input ROI is a single ROI or a vector of ROIs The locations have a one to one correspondence with the label strings in the Labels output Parameters rotation The rotation parameter controls the degree of rotational invariance allowed in the match between the model and the blobs This parameter may be used to specify rotationally Invariant or Non Invariant matching or matching over a Range of angles In the Invariant mode the operator uses a set of features that is invariant under rotation and reflection to match the blobs to the font model In this mode of operation rotated characters can be matched to model characters very quickly It is not possible to reliably distinguish between characters that are very nearly rotated or reflected versions of other characters For example the characters 6 and 9 exhibit this problem in many fonts The Non Invariant mode of operation adds rotation and reflection sensitive features to the match criteria allowing the operator to distinguish between problem characters when the approximate orientation of the characters is known The weighting of rotation sensitive features relative to the invariant features
110. only characters from the splitChars parameter string If it is set to not in set the prefix will contain only characters not in the splitChars parameter string Parameter Information splitBy Type int Default 0 Values position 0 char 1 prefix 2 splitPosition Type int Default 0 splitChars Type CorString Default splitAt Type int Default 0 Values first 0 last 1 split Type int Default 0 Values before 0 after 1 prefixChars Type int Default 0 Values in set 0 not in set 1 Demo iGraph Operators Blobs GetBlobFeatures igr Operators Blobs GetBlobFeaturesMasked igr Operators Blobs SelectFeatures igr C Prototype CorOpRtn cor_splitString CorString In CorString Prefix CorString Suffix int splitBy int splitPosition CorString splitChars int splitAt int split int prefixChars Headers include WITHOME h wConvert h Libraries WITHOME lib wObj lib WITHOME lib wConvert lib sprint Print formatted output to a string in the style of C s sprintf Description The sprint operator prints formatted output to a string in the style of C s sprintf function The operator converts the input objects to strings and forms the output string based on the format parameter The inputs must be numeric or string objects The format parameter is a sprintf style format string that specifies up to three arguments The operator scans the format string copying charac
111. or color PPM types depending on the type of the WiT image imported You can specify raw bytes or ASCII text format with the data storage parameter If you choose yes for the bitmap option the operator will convert the input WiT image to a bit mapped format by converting all non zero pixels to off and all zero pixels to on no matter what the WiT image type is For XBM format writeImage converts a WiT image to a bitmap by setting all non zero pixels to 0 and all zero pixels to 1 and writes it to disk Color is not supported for XBM format For TIFF format writeImage converts a WiT RGB image to 24 bit RGB TIFF format and converts a greyscale image to greyscale format writeImage will not use the colormap format For BMP format writeImage does not support 16 bit images The operator scales a 16 bit image to an 8 bit image before writing to disk writeImage saves a color image using 24 bit format JPEG images are written as single scan sequential JPEG files All gray scale images stored in the JPEG format are scaled to 8 bit images before compression The quality parameter controls the trade off between speed file size and compression losses Lower values give faster compression and produce smaller files but result in greater compression losses The quality values correspond to the 0 100 scale recommended by the Independent JPEG Group The raw image format is described in the readImage operator help page Color images are writ
112. output is a vector of Blob objects Each Blob object contains a single run length The run lengths are grouped into contiguous regions Within each group the run lengths are sorted by increasing line number and then by the x location of the end of the run The groups are sorted on the last run length in the group The operator keeps a running line count the run lengths and information derived from them are referenced to the beginning of the first image in the sequence of images Parameters connected The connectivity of the output blobs is specified by the connected parameter When this parameter is set to 4 pixels that share a common edge are considered to be connected so each pixel has four neighbours When it is set to 8 pixels that touch corner to corner are also considered connected so each pixel has eight neighbours type The type parameter determines how the input image will be interpreted When it is set to binary all non zero pixels are considered to be equivalent and the run lengths will represent runs of non zero pixels When it is set to labeled pixels with the same value are considered equivalent and run lengths represent runs of equal valued non zero pixels maxBlobHeight The maxBlobHeight parameter sets an upper limit on the height of blobs produced by the operator When the information derived using the webGetBlobs operator is used with images produced by the webWaterfall operator the maxBlobHeight value shoul
113. parameters control the output text If the font parameter is set to no fext no text is produced Parameter Information color Type int Default 0 Values red yellow 0 yellow green 1 aqua blue 2 white yellow 3 black blue 4 font Type int Default 0 Values helvetica 0 times 1 courier 2 no text 3 fontSize Type int Default 12 fontStyle Type int Default 0 Values normal 0 bold 1 italic 2 bold italic 3 Demo iGraph Operators Measurement CircleCenters Operators Segmentation Watershed 139 C Prototype CorOpRtn cor_plotAngle CorVector RadialPoints CorObj Vertex CorGraphicVector Out int color int font int fontSize int fontStyle Headers include WITHOME h wMeas h Libraries WITHOME lib wObj lib WITHOME lib wMeas lib plotDist Compute and plot the distances between points in a vector ed Description The plotDist operator calculates the distances between adjacent points in a vector of points and creates a vector of graphic objects which allow the distances to be displayed on an image The input may be a vector of Points Real points point type Graphics point type Geometry objects or Edge objects with a point type geometry If the input vector contains N points N floating point values are calculated for the distances in pixels between points 0 and 1 1 and 2 N 2 and N A vector of Graphics is created which includes the points and lines conn
114. random number seed is generated from a value derived from the system clock so a different random number sequence is produced each time the graph is run When the useSeed parameter is set to Yes the value specified in the seed parameter is used to initialize the random number sequence producing repeatable sequences of random numbers The seed value is a three element vector of short integers The operator s NextSeed output value is the seed value that will continue the random sequence used by the operator Parameter Information useSeed Type int Default 0 28 Values No 0 Yes 1 seed Type CorVector Default 1 0 0 type Type int Default 0 Values int 0 float 1 low Type float Default 0 high Type float Default 100 reset Type int Default 0 Values No 0 Yes 1 Demo iGraph Operators Calculate Lut C Prototype CorOpRtn cor_random CorObj Out CorUshortVector NextSeed int useSeed CorUshortVector seed int type float low float high int reset Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib sub Subtract two objects Description Subtract the second input value from the first input value See add for more information about supported types Subtracting a vector or image second input from a scalar first value is not allowed Demo iGraph Operators Calculate Lut C Prototype CorOpRtn cor_sub CorObj A
115. reference into a graphic object or vector of graphic objects in a frame of reference with a different y origin This operator is typically used to transform features produced by the webGetBlobs operator into the frame of reference of the images in the original input image stream or of images produced by the webWaterfall operator Inputs Features The input must be a Feature object or a vector of Feature objects Outputs Graphics The output is a Graphic for a single Feature input or a vector of Graphics for a vector of Features input derived from the information in the Features An output Graphic may be a point representing the blob centroid a rectangle representing the bounding box a line representing the major axis of the best fit ellipse or a polygon representing the perimeter The output graphic is offset by subtracting the amount specified in the offset parameter from the y coordinate of each point Parameters featureType The type of feature for viewing is specified by the featureType parameter and can be one of centroids ellipse axes bounding boxes or perimeters In order to plot perimeters input feature vector must have been previously passed through the webGetPerimeter operator offset The offset parameter specifies the line number of the first line in the frame of reference of the output graphics relative to the frame of reference of the input Features This parameter is typically used as an input parameter The
116. right lt lt shift left expression utilize expression to determine result The indicated operation is performed on each pixel in A The expression field allows the user to specify a C style mathematical equation which will determine the output image The image can be specified in this equation using the variable A while the constant is specified using K The result for each pixel can be affected by that pixel s particular column and row using the X and Y variables respectively For example the expression X gt 5 A gt gt K A will produce an output image where every pixel was right shifted by K bits unless the pixel fell within the first five columns of the image For more on the expression syntax and use see calc Parameter Information unaryOp Type int Default 0 Values 0 1 2 3 OR 4 AND 5 gt gt 6 lt lt 7 expr 8 expression Type CorString Default A outputType Type int Default 0 Values default 0 8 bit unsigned 1 8 bit signed 2 16 bit unsigned 3 16 bit signed 4 float 5 Demo iGraph Operators Calculate CollectAverage C Prototype CorOpRtn cor_unaryoOp CorImage A CorObj constant CorImage Out int unaryOp CorString expression int outputType Headers include WITHOME h wCalcul h 30 Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib Calibration The Calibration library supports 2D calibration which provides means to c
117. search will then only use patterns within the sampling distance of the pattern that produced the candidate match patSampling The patSampling parameter specifies the sample spacing in the pattern vector for the candidate search when applyPatterns is set to Sample The best setting for the applyPatterns and patSampling parameters depends on the structure of the variation in the pattern vector The All setting produces a more robust search but takes longer than a sampled search This mode is most appropriate when the structure of the variation within the pattern vector is unknown or unpredictable The Sample setting may often be used to speed up the search when the structure is known If the patterns are distinguishable at low resolution patSampling should be set to 1 In this case all patterns are evaluated at low resolution but only the best candidate pattern is refined If the patterns are very similar at low resolution the parameter should be set to the size of the pattern vector This will result in an initial search with one representative pattern followed by refinement with all patterns Intermediate values of this parameter are appropriate when the patterns change in some incremental way across the vector so that patterns are more like patterns nearby in the vector than they are like more distant patterns Typical examples include patterns produced by vectors of rotated images such as those produced by the patRotate o
118. specified using the outType parameter The output image is the same size as the ch0 input image and the output image ROI is the intersection of the input image ROIs Parameter Information outType Type int Default 0 Values RGB 0 HSV 1 Yuv 2 Demo iGraph Operators Color HS V C Prototype CorOpRtn cor_mergeChannels CorImage Ch0O CorImage Chi CorImage Ch2 CorImage Out int outType Headers include WITHOME h wColor h Libraries WITHOME lib wObj lib WITHOME lib wColor lib putChannel Put an image into one channel of a multichannel image EE i Description The putChannel operator inserts an image into a single channel of a multichannel image The multichannel input image must be a RGB HSV or Yuv image The inserted input image must be an 8 bit unsigned image The output image is the same type and size of as the multichannel input image The output image ROI is the intersection of the ROIs of the two input images The channel parameter specifies the channel into which the gray scale image is inserted For RGB HSV and Yuv images respectively channel 0 is the red hue or Y intensity channel channel is the green saturation or u channel and channel 2 is the blue value or v channel Parameter Information channel Type int Default 0 Values 0 0 1 1 2 2 Demo iGraph Operators Color Brighten C Prototype CorOpRtn cor_putChannel 37 CorImage In0 CorImage In
119. text The image is sampled at equally spaced intervals along the specified path The sampleSpacing parameter controls the distance between samples and bilinear interpolation is used to obtain subpixel values for the samples as in the getPath operator The smoothing parameter allows a moving average filter to be applied to the sampled path values The value of the parameter specifies the width of the filter in pixels The Edges output is a vector of edges that meet the criteria specified by the parameter settings Each element of the edge vector is a structure containing a Geom object that locates the edge a direction field that specifies the edge direction level and slope fields that indicate the edge strength a serial number and a label The edgeType parameter can be set to detect rising falling or rising or falling step edges or positive peak edges The way in which the location level and slope values are calculated depends on the type of edge The Geom element that locates the edge is an Fpoint that gives the location in pixel coordinates of the point where the edge crosses the path Edge crossing points between the sample point positions are located by linear interpolation For step edges the edge location can be chosen to be either the edge s half value point or the point of maximum slope by setting the edgeLocation parameter When this parameter is set to HalfValue the start and end of the edge are defined by where the slope c
120. the combined feature values for all the blobs represented by the input Blob feature vector The output is a single Blob feature structure that contains the size npixels field bounding box xmin xmax ymin and ymax fields and so on of the combined blobs see the getFeatures operator for a complete list of fields The output Blob feature structure cannot be used as input to the getPerimeter operator since in general the blobs represented are not connected The operator sets the perimx and perimy fields of the structure to 1 to prevent misuse of the structure Parameter Information id Type int Default 0 Demo iGraph Operators Blobs MergeFeatures C Prototype CorOpRtn cor_mergeFeatures CorFeatureVector In CorFeature Merged int id Headers include WITHOME h wBlob h Libraries WITHOME lib wObj lib WITHOME lib wBlob lib minSizeFeatures Remove small Blob features from a Blob feature vector Description The minSizeFeatures operator takes an input Blob feature vector and strips away all blobs that have fewer pixels than the number specified in the user parameter minPixels or are smaller than minWidth or minHeight The renumberlds parameter allows the user to either renumber the remaining Blob features starting at a blobs id of 1 or leave the input Blob feature id s intact Parameter Information minPixels Type int Default 1 minWidth Type int Default 1 minHeight Type int
121. the ocrRoiBin operator Parameters background The background parameter is set to light if the input image has a light background with dark text or to dark if the input image has a dark background with light text Image pixels with a value below the threshold value are set true non zero and those above the threshold as set false zero if the background has been specified as light The opposite is true if the background has been specified as dark calcThresh If the caleThresh parameter is set to No the thresholds supplied in the threshold parameter are applied to the image ROIs directly If the caleThresh parameter is set to Yes the thresholds are calculated from the pixel values in each ROI thresholds The threshold parameter is a vector of float values that set the thresholds used when the caleThresh parameter is set to No The float vector must be the same size as the ROIs vector threshFactor When the calcThresh parameter is set to Yes the threshold for each ROI is based on the mean and minimum pixel values in the ROI and is controlled by the threshFactor parameter The threshFactor parameter is usually set to a value between 0 and 1 A value of 1 will give the ROI mean as the threshold value Lower values produce thresholds nearer the foreground value in the ROI and usually result in thinning of the blobs in the binarized image Parameter Information background Type int Default 0 Values light 0 dark 1 c
122. the outputs are vectors of float and Graphic values that have a one to one correspondence to the input points If the extendLine parameter is set to No the distance and nearest point are calculated with respect to the line segment in the Line input If this parameter is set to Yes the values are calculated with respect to an infinite length line through the segment endpoints If the directional parameter is set to No the distance values are all positive values indicating the distance of the input point from the line If the directional parameter is set to Yes the distance values are positive on the right side of the input line and negative on the left Parameter Information extendLine Type int Default 0 Values No 0 Yes 1 directional Type int Default 0 Values No 0 Yes 1 Demo iGraph Operators Measurement CircleCenters Operators Segmentation Watershed C Prototype CorOpRtn cor_distPtToLine CorObj Pt CorObj Line CorObj Dist CorObj NearPt int extendLine int directional Headers include WITHOME h wMeas h Libraries WITHOME lib wObj lib WITHOME lib wMeas lib getAngle Compute the angles between points in a vector ae e j Description The getAngle operator calculates the angles between adjacent points in a vector with respect to a specified vertex point The vector of radial points is received at the top input the vertex at the bottom input The inputs may be a points
123. the setNCols parameter is set to Yes and all element 1 D objects are the same length a 2 D object is produced in which each element of the input becomes a row in the output Parameter Information setNCols Type int Default 0 Values No 0 Yes 1 C Prototype CorOpRtn cor_flattenVector CorVector In CorObj Out int setNCols Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib ge Extract an element from a vector Ji wt Description ge performs the same function as getElem ge has a small icon which makes it more convenient to use in large igraphs See Also getElem Parameter Information method Type int Default 0 Values first 0 last 1 byElement 2 element Type int Default 0 C Prototype CorOpRtn cor_getElem CorObj Vector CorObj newElement CorObj newVector int method int element 161 Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib getElem Extract an element from a vector Description The getElem operator copies an element from the input vector or linked list and outputs the extracted element and the remainder of the vector or list The element to be extracted can be specified using the method parameter to be either the first last or an intermediate element of the vector or list If the byElement option is selected then the element parameter is u
124. to use in large igraphs See Also getField Parameter Information field Type CorString Default xc C Prototype CorOpRtn cor_getField CorObj In CorObj FieldValue CorString field Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib partition Partition a vector of objects on a named field amt Description The partition operator allows the user to partition a vector of simple numeric objects or complex objects Partition of complex objects is based on the value of a particular field within the object The field must be one of the numeric or character basic types The field is specified using the field parameter If the elements of the input vector are simple numeric types the the field parameter must be left empty The vector is partitioned by comparing the field value to the value specified by the value parameter The type of comparison may be equal greater than less than not equal greater than or equal less than or equal or a range of values as specified by the condition parameter When the range comparison is used the value parameter specifies the lower bound of the range and a second parameter the hiValue parameter specifies the upper end of the range The range is inclusive so objects with field values equal to the the end point parameter values are included in the range Elements which meet the condition are collected in the TrueVecto
125. top rows of images for which the total number of rows received has not yet reached the specified height are blank zero valued Offset The Offset output is the line number of the first row of the current output image relative to the start of the first image in the sequence This value is negative for the initial image s in a sequence Parameters height The height parameter specifies the height of the output image It must be greater than the height of the images in the input image stream 288 reset When the reset function is not required the reset parameter is set to No If the reset function is required the reset parameter is used as an input parameter Sending a zero to the input parameter results in normal operation Sending a one setting the parameter to Yes will clear the previous stored images after producing the current output image The next execution of the operator will produce an image with its top rows blank The line count is also reset after the current execution Parameter Information height Type int Default 256 reset Type int Default 0 Values No 0 Yes 1 C Prototype CorOpRtn cor_webWaterfall CorImage Block CorImage Waterfall int Offset int height int reset Headers include WITHOME h wWeb h Libraries WITHOME lib wObj lib WITHOME lib wWeb lib Frame Grabber and Hardware SeriallO The SerialIO library contains operators for reading from and writing t
126. two from one to 32 rejectThresh The rejectThresh parameter is used to set a rejection threshold for match scores below which no matches will be considered The parameter must be set to a value from zero to one where zero indicates no correlation between the pattern and target and one indicates a perfect match If no part of the target image matches the pattern sufficiently well empty output vectors are produced minContrast The minContrast parameter is used to specify the minimum contrast required in the target image to produce a match The parameter specifies the minimum standard deviation of the pixel values in a possible match region relative to the standard deviation of pixel values in the pattern model image This parameter is usually set to a value between zero and one although values greater than one can be used if the target image has higher contrast than the model images maxTilesOmitted If the pattern model breaks the pattern images into multiple tiles the 263 264 maxTilesOmitted parameter allows the operator to omit some of the tiles while calculating the best matching score The operator will use the best score for a location that omits at most the number of tiles specified by this parameter The operator will always use at least one tile to calculate the best score Setting this parameter to a value greater than zero provides some tolerance to partially occluded or damaged target regions The MultipleMatches
127. values If Separate is selected subsequent plots are displayed on a separate graph window If Replace is chosen each subsequent plot clears the existing plots on the graph window before being plotted Parameter Information name Type CorString Default iMeas type Type int 61 Default 0 Values point 0 line 1 polyline 2 polygon 3 rectangle 4 circle 5 UnitName Type CorString Default pixel UnitsPerPixel Type float Default 1 Placement Type int Default 0 Values bitwise OR Fix position 1 X Type int Default 0 Y Type int Default 0 iThresh Interactive thresholding 5i aly Description The iThresh operator displays a grayscale image in a window and provides controls for selecting the threshold level The effect of the threshold level can be seen immediately as the level is moved When the desired threshold effect is achieved click OK and the image you see is sent to the output lE E lx irri Beatie fo ieee Balen erm r ke Chew pripe fe og f Properties Main panel 62 grTitle The title of the graph If there are multiple graph operators in an igraph you can send all the plots to the same graph by setting the grTitle of the operators to the same name Also if you have set Auto clear in the Options panel to false then you can plot curves from different igraphs to the same graph provided that the grTitle are set to the same name
128. wFit h Libraries WITHOME lib wObj lib WITHOME lib wFit lib 110 pix2pt Convert pixel locations to points a F Description The pix2pt operator will take a binary image and produce a point vector describing the point locations of every non zero pixel in the image ROI The vector is limited in size by the maxPts parameter Parameter Information maxPts Type int Default 100 Demo iGraph Operators Geometry convexHull igr C Prototype CorOpRtn cor_pix2pt CorImage In CorPointVector Points int maxPts Headers include WITHOME h wFit h Libraries WITHOME lib wObj lib WITHOME lib wFit lib reduceLines Find the lines that best fit the sequence of the points Description The reduceLines operator produces approximations to polygons or polylines that contain fewer line segments than the originals The input may be a polygon or polyline Geometry or Graphic object a vector of polyline or polygon Geometry or Graphic objects or a vector of Point or Fpoint objects A vector of points is treated as if the points were endpoints in a single polyline The output is the same type as the input except for the point vector input which produces a polyline Geometry object The polylines and polygons are approximated using a successive splitting approach For a polyline the initial approximation is a single segment polyline connecting the endpoints of the original polyline If the original polyli
129. width of perimeter points Point vector containing all perimeter points Parameters yOffset The yOffset parameter specifies the line number of the first line in the input image relative to the frame of reference of the input Features This parameter is typically used as an input parameter The line number offset outputs of operators such as webWaterfall and webLineCounter may be used to provide the input value connected The connected parameter can be used to specify whether the perimeter pixels should be eight connected where diagonal neighbours are considered to be touching or four connected where they are not type The type parameter is used to to specify whether the input image is considered a binary image in which case non zero regions will be traced or a labeled image in which case equal valued non zero regions will be traced The connected and type parameters should be set to the same values as the connected and type parameters on the webGetBlobs operator that was used to generate the input feature vector If they are not the operator may have trouble correctly tracing the perimeter Parameter Information yOffset Type int Default 0 connected Type int Default 1 Values 4 0 8 1 type Type int Default 0 Values binary 0 labeled 1 C Prototype CorOpRtn cor_webGetPerimeter CorFeatureVector In CorImage BinaryImage CorFeatureVector Out int yOffset int connected int type Heade
130. 0 maxTilesOmitted Type int Default 0 localSearch Type int subPixel Default 0 Values Fast 0 All 1 Type int Default 0 Values No 0 Yes 1 resolution Type int scale Default 3 Values 2 0 4 1 8 2 16 3 32 4 64 5 Type int Default 2 Values 2 0 4 1 8 2 16 3 32 4 maxCandidates Type int Default 4 candidateThreshold Type float Default 0 5 candidateRelThresh Type float Default 0 applyPatterns Type int Default 0 Values All 0 Sample 1 patSampling Type int Default 1 output Type int Default 0 Values Final Values 0 Candidates 1 maxMatches Type int Default 1 relThresh Type float Default 0 xVicinity Type int Default 0 yVicinity Type int Default 0 C Prototype CorOpRtn cor_patFind1 CorImage Target CorGraphic ROI CorPatModell PatModel CorFpointVector Locations CorFloatVector Scores CorIntVector Indexes float rejectThresh float acceptThresh float minContrast int maxTilesOmitted int localSearch int subPixel int resolution int scale int maxCandidates float candidateThreshold float candidateRelThresh int applyPatterns int patSampling int output int maxMatches float relThresh int xVicinity int yVicinity 251 Headers include WITHOME h wSearch h Libraries WITHOME lib wObj lib WITHOME lib wSearch lib patFind2D1 Find one of a 2D array of patterns Be Description The pa
131. 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 298 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 Demo iGraph Operators Video Acquisition trigger C Prototype CorOpRtn acqLut int lutNumber CorIntVector lut Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqOverlay Overlay image or graphics on video window nt a i Description The acqOverlay operator is available on frame grabbers which support live video display This operator takes an image or Graphic vector as input and draws the data to the live video display window at the specified x and y offset Typically the acqLive operator is used first to bring up the live video window then the acqOverlay operator is used to display results The acqOverlay operator can only handle Graphic objects of type point circle rectangle text and l
132. CIE Yuv color space and the match score is calculated using the gray scale pixel values The other color space modes are not yet implemented If the model image is not a color image the parameter value is is ignored and the color mode is automatically set to brightness Parameter Information subPixel Type int Default 0 Values No 0 Yes 1 mask Type int Default 0 Values No 0 Yes 1 tileCols Type int Default 1 tileRows Type int Default 1 colorMode Type int Default 0 Values brightness 0 fixColor 1 varyColor 2 C Prototype CorOpRtn cor_patInit1 CorObj ModelImages CorObj MaskImages CorPatModell PatModel int subPixel int mask int tileCols int tileRows int colorMode Headers include WITHOME h wSearch h Libraries WITHOME lib wObj lib WITHOME lib wSearch lib patinit2D1 Initialize 2D array of patterns i Description The patInit2D1 operator prepares an array of pattern models from a vector of image vectors for use with the patFind2D1 operator These operators facilitate searching for patterns with two distinct types of variation for example rotated and scaled versions of a single pattern or rotated versions of each of several different patterns The pattern model images may be broken up into equal sized rectangular subimages or tiles The pattern matching operators can optionally use a subset of the tiles when calculating match scores to improv
133. CORECO IMAGING Coreco Imaging 7075 Place Robert Joncas Suite 142 St Laurent Quebec Canada e H4M 222 http www imaging com WiT Operator Reference Manual Edition 8 0 Part number OC WITM OPREF NOTICE 2003 Coreco Imaging Inc All rights reserved This document may not be reproduced nor transmitted in any form or by any means either electronic or mechanical without the express written permission of Coreco Imaging Inc Every effort is made to ensure the information in this manual is accurate and reliable Use of the products described herein is understood to be at the user s risk Coreco Imaging Inc assumes no liability whatsoever for the use of the products detailed in this document and reserves the right to make changes in specifications at any time and without notice Microsoft and MS DOS are registered trademarks Windows Windows 95 Windows NT and Windows XP are trademarks of Microsoft Corporation All other trademarks or intellectual property mentioned herein belong to their respective owners Printed on September 30 2003 Document Number OC WITM OPREF Printed in Canada CONTENTS CONTENTS sssscssssscsscsssocstessessetsesssonseosseasensevesoonsssconscoces i CALIDVAtC CAMEL ccesccssescesseeseceesesessseeseesessenseees 30 Stannard siini hesan sa e eE iS 1 calibrateDefauit EA edie 3l calibrateRal oneen narre nE ORS 32 Blobs san a Se Re 1 calibrateScale n sssssssssssssss
134. Complex h Libraries WITHOME lib wObj lib WITHOME lib wComplex lib Convert Type The Convert Type library supports changes in object type and data representation The library includes operators that cast between numeric types string management operators and operators that prepare objects for display cast Convert basic object to another type large icon Description The cast operator converts its input object to the output type specified by the type parameter The operator is primarily used to convert between the various integer and float object types including conversions between Point and Real Point objects or to convert numbers to strings or strings to 40 numbers It can be used to cast single objects or vectors of objects Conversions between floating point objects float and double types integer objects byte ubyte short ushort int uint hex8 hex16 hex and char types and between Point and Real Point objects result in the value of the input object being assigned to an output object of the selected type If the value of the input object falls outside the range of the output object type the value is clipped to the range If the input is a vector of one of these types the output will be a vector of the same size in which the elements are of the type specified in the type parameter The conversion parameter determines how float or double types are converted to one of the integer output types and h
135. Default 1 renumberlds Type int Default 0 Values No 0 Yes 1 Demo iGraph Operators Blobs MinSizeFeatures C Prototype CorOpRtn cor_minSizeFeatures CorFeatureVector In CorFeatureVector Out int minPixels int minWidth int minHeight int renumberIds Headers include WITHOME h wBlob h Libraries WITHOME lib wObj lib WITHOME lib wBlob lib selectBlobs Select blobs by blob Id si oon Description The selectBlobs operator partitions a blob vector based on the blob id number The input is a Blob vector a run length encoded representation of a binary of labeled image produced by the getBlobFeatures operator Blobs or contiguous regions are represented in the Blob vector by run lengths with the same id number Run lengths belonging to the same blob are in scan line order ie sorted by increasing row in the vector Blobs are selected by specifying an id number or numbers Run lengths with ids matching the specified ids are output at the as a Blob vector at the Selected output all run lengths that do not match one of the specified ids are output at the NotSelected output The ids are specified using the select parameter The parameter may be any scalar value or vector of scalar values in which case the values are used directly to specify the ids or it may be a Feature or vector of Features in which case the Feature id value is used The Feature vector is also produced by the getBlobFeatu
136. Description The convexHull operator calculates the convex hull or envelope of a binary image set of points or geometric object An image input must be unsigned 8 or 16 bit image The convex hull encloses all non zero pixels in the image ROI based on the pixel centers The operator also accepts vectors of Point or Real point objects and both vectors of geometric objects and single geometric objects Geometric objects include Graphic Geometry Edge and Line objects The operator output is a polygon Graphic object The color associated with the graphic can be set using the parameters on the Color subpanel Parameter Information outline Type int Default 1 Values off 0 on 1 penR Type int Default 255 Values 0 255 penG Type int Default 0 Values 0 255 penB Type int Default 0 Values 0 255 fill Type int Default 0 Values off 0 on 1 fillR Type int Default 0 Values 0 255 fillG Type int Default 0 Values 0 255 fillB Type int Default 0 Values 0 255 Demo iGraph Operators Geometry convexHull igr C Prototype CorOpRtn cor_convexHul11 CorObj In CorGraphic Hull int outline int penR int penG int penB int fill int fil1R int fillG int fi11B Headers include WITHOME h wFit h Libraries WITHOME lib wObj lib WITHOME lib wFit lib fitCircleToPts Iterative least squares circle fit to points Description The fitCircleToPts op
137. DoubleThresh C Prototype CorOpRtn cor_thresh_1 CorImage In CorImage Out float lo float hi int invert int levels int result CorObj semiFillColor CorOpRtn cor_thresh_1_consume CorImage In float lo float hi int invert 196 int levels int result CorObj semiFillColor Headers include WITHOME h wSegment h Libraries WITHOME lib wObj lib WITHOME lib wSegment lib watershed Perform a watershed image segmentation 1 Tt 4 daii Description The watershed operator segments an integer or floating point image into regions analogous to catchment basins in topographic elevation maps using the algorithm of Vincent and Soille PAMI 13 6 583 597 1991 Each region corresponds to a local image minimum and the surrounding basin pixels for which there is a descending path to that minimum and to no other local minimum Pixels that have a descending path to more than one local minimum are not included in any region except in the case of plateaus Plateaus are flat areas of more than one pixel width between basins these regions are divided according to their distance from the adjoining basins with plateau pixels assigned to the region associated with the nearest basin Pixels mid way between two basins may not be assigned to a region The operator output is an unsigned 16 bit image in which the pixels of each region are given a distinct value or label from one to the nu
138. E h wWeb h Libraries WITHOME lib wObj lib WITHOME lib wWeb lib webRealign Align images from multiple line scan cameras Description The webRealign operator corrects alignment errors in a sequence of images created by horizontally concatenating images produced by a number of side by side cameras The images in the sequence represent vertically adjacent blocks in the imaged space The two alignment errors handled by the operator are overlap in field of horizontal field of view between adjacent cameras and vertical misalignment of the cameras The non overlapping columns from each camera and the vertical offset associated with each camera are specified using the operators parameters The operator uses rows of previous images in the stream to correctly align pixels in the rows of the offset portions of the images Inputs Outputs In The input image is assumed to be a composite image in which the images from several side by side cameras are concatenated into a single wide image The image width must be a multiple of the number of cameras specified implicitly in the operator parameters All images in a stream of images received by an instance of the operator must be the same size and type Out The output image is formed by concatenating the portions of the input image specified in the validRange parameter vector These portions are offset vertically by the amounts specified in the offset parameter vector The
139. E lib wSystem lib 209 rdText Read a text file and convert to WiT data objects cat a Description Read a text file and convert to WiT data objects Parameter Information filename Type CorFile Default test txt objType Type int Default 0 Values char 0 int8 1 int16 2 int32 3 long 4 uint8 5 uint16 6 uint32 7 ulong 8 hex8 9 hex16 10 hex 11 float 12 double 13 String 14 array Type int Default 0 Values as image 0 as vector of vectors 1 multiLineStrings Type int Default 0 Values delimited 0 whole file is one string 1 C Prototype CorOpRtn cor_rdText CorObj Out CorFile filename int objType int array int multiLineStrings Headers include WITHOME h wSystem h Libraries WITHOME lib wObj lib WITHOME lib wSystem lib 210 readimage Read standard image format files Description The readImage operator reads an image file in one of a number of common image formats including the WiT object format or a raw image data file from disk The filename parameter is the name of the file The object search path is used to search for the specified file The type parameter specifies the image format Currently the following formats are supported Graphics Interchange Format 8 rx frci no _ supported pcx e PaintBrush 8 PC PaintBrush 8 bi Portable Bitmap Format PBM Soar oa SUN Sun Raster File Format Raster File Format
140. E lib wSystem lib ringBell Generate a beep i Description The ringBell operator sounds the system warning beep when an object is received at the input Parameter Information bellType Type int Default 0 Values Sharp 0 Low 1 C Prototype CorOpRtn cor_ringBell int bellType Headers include WITHOME h wSystem h Libraries WITHOME lib wObj lib WITHOME lib wSystem lib splitName Split a string into path base and extension m Eis Description The splitName operator takes a string filename as input The filename should be an absolute path for a file in the form path base ext The Path component is everything in front of the last slash both and are accepted If filename does not contain a slash Path is set to NULL The Base component is everything after the last slash but in front of the last period and Ext is everything after the last period If there is no period Ext is set to NULL C Prototype CorOpRtn cor_splitName CorString Filename CorString Path CorString Base CorString Ext Headers include WITHOME h wSystem h Libraries WITHOME lib wObj lib WITHOME lib wSystem lib stopSound Stop playing a waveform sound file ah fr Description Stop playing a waveform sound file C Prototype CorOpRtn cor_stopSound Headers include WITHOME h wSystem h Libraries WITHOME lib wObj lib WITHOME lib wSystem lib st
141. Feature structures produced by the getFeatures operator using the field parameter npixels This syntax applies only to top level vectors and cannot be applied to vectors that are themselves fields of the input object It is possible for example to form a vector of x coordinates from the first point in a vector of Graphic objects using the field name geom plist data 0 x It is not possible however to extract all of the x coordinates of the points in a Graphic object using this syntax since the plist vector is not at the top level In this case two getField operators must be used First the plist field must be extracted using the field name geom plist The x coordinates can then be extracted from this vector with the x field name Parameter Information field Type CorString Default xc C Prototype CorOpRtn cor_getField CorObj In CorObj FieldValue CorString field Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib getRect Extract image bounding box pia Description The getRect operator creates a graphic object describing a rectangle with the same dimension as the input image The rectangle has a top left corner of 0 0 and a bottom right corner of W 1 H 1 where W and H are the width and height respectively of the input image C Prototype CorOpRtn cor_getRect CorImage In CorGraphic Rect Headers include WITHOME h wObjman
142. For general objects the size is just big enough to show the entire object with the default font If the object size is larger than the screen then the window size is limited to the screen size and scroll bars appear automatically You can also specify a maximum size with the MaxW and MaxH parameters in the Placement sub panel MaxW and MaxH only affects the initial window size they do not limit any subsequent manual resizing of the object window More controls are available from the pop up menu press right mouse button inside window Pop up menus are different for image and non image data objects Display is a powerful and often used tool Refer to the section Operators Libraries in the User s Manual for details See Also smallDisplay Parameter Information name Type CorString Default display Placement Type int Default 0 Values bitwise OR Fix position 1 Max size 2 Min size 4 X Type int Default 0 If Fix position of the Placement parameter is enabled bit 1 0x1 the X parameter specifies the horizontal position on the screen of the upper left hand corner of the display window If Fix position is not enabled the X parameter is ignored Y Type int Default 0 If Fix position of the Placement parameter is enabled bit 1 0x1 the Y parameter specifies the vertical position on the screen of 53 Maxw MaxH MinW MinH Range RangeLow RangeHigh Equalize EqLow EqHigh
143. For signed input images the pixel values are offset to produce index values that begin at zero For example in a signed byte image a pixel value of 128 the minimum pixel value indexes element 0 of the LUT a value of 127 indexes element 1 and so on When the LUT is a signed int vector the output image is the same type as the input image The pixel values are mapped to the LUT elements in the same way as with other input image types It is the user s responsibility to ensure that the LUT element values are not outside the range of the output image type Demo iGraph Operators Calculate Lut C Prototype CorOpRtn cor_applyLut CorImage In CorVector Lut CorImage Out Headers include WITHOME h wCalcul h 13 Libraries calc WITHOME lib wObj lib F Perform arithmetic and logic operations on an WITHOME lib wCalcul lib object autoEqualize ai aji lk ESj Automatic linear rescaling of image values Description 2 yal t F Same as calc but with only one input Description Parameter Information expression Type CorString The autoEqualize operator performs automatic Default A image contrast enhancement by computing an equalization interval based on the input integer userConstant Type double image intensity frequency distribution The interval Default 0 is determined by searching the image histogram type Type int from each end until two values are found for which Default 0 a specified pe
144. ITHOME lib wObj lib WITHOME lib wSegment lib kMeans Automatically segment an image by pixel value Description The kMeans operator partitions an image into k clusters based on the pixel value In the resulting partition each pixel is closer in value to the mean value of the pixels in the cluster to which it belongs than it is to the mean of any other cluster In general there is not a unique partition that satisfies this condition so the result may depend on choice of initial mean values for the iterative method The input to the operator is an integer or float image The output image at the top output port is an unsigned 8 bit segmented image with each pixel 191 assigned the label associated with its cluster The labels are numbers from 0 to k 1 ordered such that 0 represents the lowest value range and k the highest value range The other three outputs are vectors of k numbers representing the mean variance and number of pixels for each cluster The initMeans parameter is used to select from the two methods provided for specifying initial mean values If auto is selected then the number of clusters k must be specified with the numClusters parameter The operator will then automatically spread the initial means evenly over the range of values in the image If array is selected the initial mean values are specified by the meanValues array parameter In this case the number of clusters k is set to the array size
145. Image Var int output int width int height Headers include WITHOME h wEFilter h Libraries WITHOME lib wObj lib WITHOME lib wFilter lib Find The Find library provides operators that search images for pixel value peaks or pixel patterns The library includes template matching operators operators that search for peaks in an image and image to image correlation operators getMatch Directed search for image maxima or minima Description The getMatch operator performs directed search for values that exceed a specified threshold The operator searches specified ROIs in the target for pixels which are greater or less than a threshold value The output is a vector of graphic objects that specify locations of such pixels The operator requires two inputs the top input is an integer or float image and the bottom input is a graphic object or vector of graphic objects that controls the search strategy The operator searches the pixels included in the graphic objects specified by the input In addition point graphic objects initiate a spiral search starting at the point and extending out to the number of pixels specified by the searchRange parameter These pixels are compared to the value specified in the threshold parameter to find those pixels which are greater than the threshold if find is set to maxima or less than the threshold if find is set to minima The operator selects from among the pixels exceeding
146. Local1 Find a pattern in point neighbourhoods a ed 259 Description The patLocal operator locates the best match or matches to one or more patterns in the neighbourhood of specified points in the target image Matches between the patterns and the image are evaluated by a statistical technique using pixel grayscale values The inputs to the operator are a gray scale target image a vector of points and a vector of pattern models which may use tiled pattern images and or binary masks The size of the neighbourhood is specified by a parameter The operator outputs the location match score and pattern element number of the match or matches meeting criteria specified by the operator s parameters Inputs Target The target image can be any gray scale WiT image an unsigned or signed eight or sixteen bit integer image or a float image The operator does not accept color or complex images ROI The ROI is a vector of Real Points The operator searches the neighbourhood surrounding each point in the image The size of the neighbourhood is specified by the range parameter PatModel The pattern model is produced by the patInit1 operator from a pattern image or vector of images The pattern may be broken into equal sized rectangular tiles and or may incorporate an arbitrary binary mask as described in the patInit1 operator documentation Outputs Locations The match locations are produced as a vector of Real points giving the x
147. ME h wCalibration h Libraries WITHOME lib wObj lib WITHOME lib wCalibration lib restoreCoordinates Apply calibration transformation to geometry or graphic objects oe h Description The restoreCoordinates operator transforms coordinates in the input data object using the calibrateInfo object created by calibrateCamera operator If mode is forward then the transformation is from screen to real coordinates Otherwise the transformation is from real to screen coordinates The input may be a Geometry or Graphic object a vector of Geometry or Graphic objects or a vector of Point or Fpoint objects Parameter Information mode Type int Default 0 Values forward 0 backward 1 Demo iGraph Operators Calibration calibrate igr C Prototype CorOpRtn cor_restoreCoordinates CorObj coordData CorCalibrateInfo calibrateInfo CorObj transformedData int mode Headers include WITHOME h wCalibration h Libraries WITHOME lib wObj lib WITHOME lib wCalibration lib restorelmage Apply calibration transformation to an image a ax Description The restoreImage operator transforms the input image using calibrateInfo object created by calibrateCamera operator If mode is forward then the transformation is from screen to real coordinates Otherwise the transformation is from real to screen coordinates Parameter Information mode Type int 33 Default 0 Values forward 0 backward 1 De
148. N ceed tebek Heseson 151 zerolmage 182 a E ea e aea Pyramids 183 butt 183 DUTEROC OM sarn Ea E i S 183 burtReeStep ots bet ea eee 184 burtStep wl decimate ae east fda ice e AEE e 185 CV GTOPIUG sonriente ri a elma tln eaeia tence totoeen tice etek tees aoc crObject 2 y ALE E E EEE E E 159 showPyramid resien iee r a AE e RER 188 CHOGLER CCL a E EEN ENES 159 UN DECIMAL Geis nena Ea EEES 189 Segmentation norin aa aR AER R 189 v tBarcod maien a eink Paes a 221 adapt Ihreshanssnnniseenc nnn 189 BCIC CA E E E eistestitadebtactecsesents 221 color Thresh eedan ana aea 190 betCodabat sieeeseevraiai neaei feyeetecass 222 kMeans localAdaptThresh ccccscsccscseseesceseeeeseeseeseees 192 ELC ODES Dien soe ea aeie EAEEREN 224 LOCALP COKS ieena E 194 BOTEAIN AS ceccciccencevecsnsinteasteissynstaneestecisonaparcousunes 225 LOCATE Ih a esis eivce cscs ais eadi ls ca E sence 194 BCIEANG eaen ai Mises desks deseo 225 HAT OS 3 cas feeck sesepeavevestevidl tupserteachi sith EE able 195 BOLE OC C200 or saics ses iesenacsecashies i a 226 Wer shed i oie te ii biel Abt ei iene 196 BOLE CO20O NA oie r o EES A ARE REETA 227 LOT OX a A EER RA TE E A AE AS 197 Beth C C200 2s r ARA A EAE 227 Statisties sg i nha ede cee eee 198 PTE cccicea E A E oh tatease vas pede tes a 228 backend viele cies et A BEER se oes 198 BCLPOSTNET i e EE 228 DOUNE o ee an e alias Mae sone E AAR erh 198 BEtPNAVMACOE eceeceeeesecesereeseeseetee
149. OME lib wMatrix lib warpMatrix Finds and unwarps ECC 200 2D barcode amiy IEE Description The warpMatrix operator finds and extract images of the ECC 200 barcodes If a barcode is found then it is extracted and warped to the rectangular image If several barcodes were found then several images are placed in the output vector Inputs source The input image must be an unsigned 8 bit grayscale image with the barcode represented as white bars or dots on a black background The ECC 200 codes may be rotated scaled or warped binary The input image must be an unsigned 8 bit binary image produced from the image mentioned above Outputs unwarped The unwarped output is a vector of images Info The Info output provides a vector of graphics corresponding to each suspicious region found in the original image warpXform The warpXform output is a transformation matrix for the found ECC200 barcode Parameters cellSize The cellSize parameter determines the size of each cell comprising the 2D bar code given in pixels This size must be bigger than the distance between two cells or dots in a barcode but smaller than the silent zone around the perimeter of the barcode Parameter Information cellSize Type int Default 10 Demo iGraph Operators Matrix UPSe igr C Prototype CorOpRtn cor_warpMatrix CorImage Source CorImage Binary CorImageVector Unwarped CorGraphicVector Info CorVector WarpXform int cellSiz
150. Obj lib WITHOME lib wConvert lib splitString Split a string into two substrings vais hp Description The splitString operator partitions an input string into two substrings The point at which the string is split may be specified by a position in the string a specific character or at the end of prefix containing specified characters The method of locating the split point is specified using the splitBy parameter When splitBy position is selected the first substring contains the number of characters given in the splitPosition parameter The second substring 48 contains the remaining characters of the input string If the splitPosition value is greater than the length of the input string the first output string contains the entire input string and the second output string is empty Setting splitBy to char causes the string to be split at either the first or last as specified by the splitAt parameter occurence of any character in the splitChars parameter string The string can be split before or after the character as specified by the split parameter If splitBy is set to prefix the first output string contains a prefix of the input string made up of characters from a specified set The second output string starts at the first occurence of a character not in the set and all subsequent characters in the string whether they are in the set or not If the prefixChars parameter is set to in set the prefix will contain
151. Obj lib WITHOME lib wMeas lib distMap Produce a distance map from a binary image ie Description The distMap operator converts a binary image to a distance map image in which a pixel s value represents the distance of the pixel from object boundaries in the image The method parameter is used to specify the type of distance measure used by the operator Two approximations to Euclidean distance are available as well as two integer methods The Euclidean 3 x 3 and Euclidean 5 x 5 methods use floating point masks to accurately determine the distance between pixels in 3x3 or 5x5 neighbourhoods which are accumulated to approximate the Euclidean distance over larger regions The 3x3 mask limits errors to under 8 0 the 5x5 mask to under 2 4 In the City Block method the distance is determined to the nearest edge in horizontal and vertical steps In the Chess Board method diagonal steps are also 136 allowed and all steps are considered to be of unit length The distFrom parameter is used to specify whether the distance to each pixel is calculated from the nearest background pixel or the nearest foreground pixel In the first case all output image pixels corresponding to background zero valued pixels in the input image have a distance value of zero and output image pixels corresponding to foreground non zero valued pixels in the input image contain the distance to the nearest background pixel In the second case the for
152. On receiving a token at its input the clearDisplay operator closes all of the pop up displays that were were previously opened to display data display Display object data in a pop up window Description The display operator creates a pop up frame containing the data of the object received at its input port If the object is an image then the pop up will present a rendered image If the object is not an image then display will pop up a frame with text showing all the pertinent data contained within the object al Gl ml Ae 1 1 1 LD We 1 i 1 i r 1 Se aiia String Hello world When data arrives at a display operator the data is shown in a detached window Display detects the type of the input object If it is an image or a vector of images then it brings up an image window otherwise it brings up a general object window The position of the pop up window is controlled by the Placement option in the Advanced Options panel The default is to tile windows on the screen You can override auto placement by specifying an explicit position for the X and Y parameters in the Placement sub panel of the display property panel The name of the pop up display can be set using the name parameter Operators with the same name will display objects in the same pop up window with the new object replacing any existing displayed object For images the size of the pop up window is the actual size of the image
153. Parameter Information minScale maxScale scaleIncr minAngle maxAngle angleIncr width height Type float Default 1 Type float Default 1 Type float Default 0 1 Type float Default 0 Type float Default 0 Type float Default 1 Type int Default 0 Type int Default 0 C Prototype CorOpRtn cor_patScaleRotate CorImage In CorVector Transformed CorFloatVector Params float minScale float maxScale float scaleIncr float minAngle float maxAngle float angleIncr int width int height Headers include WITHOME h wSearch h Libraries WITHOME lib wObj lib WITHOME lib wSearch lib searchEdit Edit search parameters Description Edit search parameters Parameter Information Placement Type int Default 0 Values bitwise OR Fix position 1 X Type int Default 0 Y Type int Default 0 C Prototype CorOpRtn cor_searchEdit CorObj In CorSearchInfo Out int Placement int X int Y Headers include WITHOME h wSearch h Libraries WITHOME lib wObj lib WITHOME lib wSearch lib searchRun Search for patterns in an image Description See User s Manual for details C Prototype CorOpRtn cor_searchRun1 CorImage Image CorSearchInfo Info CorFpointVector Locations CorFloatVector Scores CorFloatVector Scales CorFloatVector Angles CorGraphicVector Graphics Headers include WITHOME h wSearch h
154. Perimeter operator The color of the output graphics can be controlled using parameters on the Color subpanel If the colors parameter is set to custom the remaining parameters can be used to specify whether the output graphics are outlines or filled graphic objects and to specify the color of the objects The default graphics are yellow outlines The label field of each graphic object is set to the id number of the associated element of the Blob feature vector Parameter Information featureType Type int Default 0 Values centroids 0 major axis 1 bbox 2 perimeter 3 bbox center 4 colors Type int Default 0 Values default 0 custom 1 outline Type int Default 0 Values off 0 on 1 penR Type int Default 0 Values 0 255 penG Type int Default 0 Values 0 255 penB Type int Default 0 Values 0 255 fill Type int Default 0 Values off 0 on 1 fillR Type int Default 0 Values 0 255 fillG Type int Default 0 Values 0 255 fillB Type int Default 0 Values 0 255 Demo iGraph Operators Blobs GetBlobFeatures igr Operators Blobs GetBlobFeaturesMasked igr Operators Blobs SelectFeatures igr C Prototype CorOpRtn cor_plotFeatures CorObj Features CorObj PlottedFeatures int featureType int colors int outline int penR int penG int penB int fill int fi11R int fillG int fi1l1B Headers include WITHOME h wConvert h Libraries
155. Prototype CorOpRtn cor_webGetBlobFeatsRoi CorImage In CorObj ROI CorFeatureVector BlobFeatures CorBlobVector Blobs int type int connected int maxBlobHeight int reset int resetBlobld int minPixels int minWidth int minHeight Headers include WITHOME h wWeb h Libraries WITHOME lib wObj lib WITHOME lib wWeb lib webGetBlobFeatures Get Blobs and Blob features from a stream of line scan images F E TETE Description The webGetBlobFeatures operator performs pixel connectivity analysis on a stream of binary or labeled images The operator extracts run lengths from the image and outputs both a set of features describing each contiguous region and the run lengths organized into blobs in the form of a Blob 274 vector Images in the sequence are treated as if the last line of the previous image is adjacent to the first line of the current image and the first line of the next image is adjacent to the last line of the current image allowing blobs to extend across more than one image Blobs are output when they do not reach the bottom of the current image and so cannot extend into the next image in the sequence Blobs that reach the bottom row of the current image are usually saved by the operator to see if they extend into the next image in the sequence The operator can be forced to output these blobs using the reset parameter The operator is approximately equivalent to applying the getBlobFea
156. Real points point type Graphics point type Geometry objects or Edge objects with a point type geometry If the input vector contains N points the output is a vector of N floating point values representing the angle in degrees between the lines connecting the vertex to points 0 and 1 to 1 and 2 to N 2 and N J A zero size vector is produced if the input vector contains fewer than two elements 137 Demo iGraph Operators Measurement CircleCenters Operators Segmentation Watershed C Prototype CorOpRtn cor_getAngle CorVector RadialPoints CorObj Vertex CorFloatVector Angles Headers include WITHOME h wMeas h Libraries WITHOME lib wObj lib WITHOME lib wMeas lib getDist Compute the distances between points in a vector Description The getDist operator calculates the distances between adjacent points in a vector of points The input may be a vector of Points Real points point type Graphics point type Geometry or Edge objects with a point type geometry If the input vector contains N points the output is a vector of N 1 floating point values representing the distance in pixels between points 0 and 1 1 and 2 N 2 and N 1 A zero size vector is produced if the input vector contains fewer than two elements Demo iGraph Operators Measurement CircleCenters Operators Segmentation Watershed 138 C Prototype CorOpRtn cor_getDist CorVector Points CorFloatVector Distances
157. Rtn cor_fht CorImage In CorImage Out int dir Headers include WITHOME h wXform h Libraries WITHOME lib wObj lib WITHOME lib wXform lib 219 SmartSeries SmartMatrix One and two dimensional bar code reading getBC412 Finds and decodes BC412 bar code z ty rn a ENNIS Description The getBC412 operator accepts a binary or grayscale image and one or more rectangular regions of interest ROIs then searches the ROIs for BC412 barcodes If a barcode is found in a region it is decoded to a string Inputs Target The input image must be an unsigned 8 bit grayscale image with the barcode represented as light bars on a dark background Roi A single ROI may be specified by a rectangular Graphic or Geometry object Multiple ROIs can be specified by a vector of Graphic or Geometry objects Outputs Decoded If the ROI is a single object then the output is a String object If a valid barcode is found the output string contains the characters represented by the barcode including the check character If no valid barcode is found an empty string is output 220 If the ROI is a vector of rectangles then the operator seeks for a barcode in each ROI The output is a vector of String objects If a valid barcode is found in a ROI the corresponding element of the output vector contains the decoded string If no valid barcode is found in a ROI the corresponding output vector element is the empty str
158. Size cancel ok pixVal reset and sample See the WiT Professional Toolkit Manual for details Parameter Information name Type CorString Default data magnification Type int Default 0 Values 1 0 2 1 4 2 C Prototype CorOpRtn cor_edit CorObj In CorObj Out CorString name int magnification Headers include WITHOME h wDisplay h Libraries WITHOME lib wObj lib WITHOME lib wDisplay lib getData Interactive specification of graphical objects Fida E a P Ginis lr Description The getData operator allows interactive specification of Graphic objects associated with an image The Graphic objects can be geometric shapes such as points lines or polygons or text A color can be associated with each object and the method of drawing geometric shapes outlined filled or both can be specified The resulting graphical objects can be displayed on the original image or other images using the overlayData operator or used in later processing and analysis steps for example with the extract or statsRoi operators 55 The getData operator is interactive requiring direct user input When an image object arrives at the input port it is displayed and the user can define graphical objects on the image using the mouse and a graphical editor panel as described below Once the objects are defined the user selects OK from the pull down menu obtained by clicking the right mouse button in
159. Sub Panel The MultipleMatches sub panel includes parameters that control the output when multiple matches are required max Matches The maxMatches parameter sets the maximum number of matches that will be produced by the operator The operator may produce fewer matches relThresh The relThresh parameter is used to set a relative rejection threshold for the matches The relative rejection threshold is determined by multiplying the parameter value by the match score of the best match found This value is then applied like the normal rejection threshold selectMode The selectMode parameter controls how the operator treats multiple matches that may correspond to the same target When this parameter is set to Location any matches found within the range specified by the xVicinity and yVicinity parameters are considered duplicates When it is set to Pattern any matches that correspond to the same element in the pattern model vector are considered duplicates When it is set to LocAndPat matches that are within the vicinity and correspond to the same element in the pattern vector are considered duplicates Only the best match of a set of duplicate matches is kept xVicinity and yVicinity These parameters specify a minimum region around a match location in which matches with lower match score will be suppressed when the location is being used to select matches Two matches will never be produced that are less than both xVicinity pix
160. T will shift the image origin to the centre of the image for presentation by the spectrum operator Parameter Information dir Type int Default 0 Values forward 0 reverse 1 output Type int Default 0 Values Default 0 User 1 outType Type int Default 0 Values Complex 0 Real 1 Imaginary 2 Magnitude 3 Phase 4 Demo iGraph Operators Transforms Transforms Operators Transforms FourierFilter Operators Transforms CrossCorrelation C Prototype CorOpRtn cor_fft CorImage In CorImage Out int dir int output int outType Headers include WITHOME h wXform h 217 Libraries WITHOME lib wObj lib WITHOME lib wXform lib fht Fast Hadamard transform an image Description The fht operator computes the Fast Hadamard transform of the input image A forward or inverse FHT is specified by the dir parameter A forward FHT expects either an integer or float input image and inverse FHT expects a float input image with both x and y dimensions a power of two The output in either case is a float image The size of the output image is equal to that of the input image regardless of the input image ROI size rounded up to the nearest power of two in each dimension Only pixels in the input image ROI are used in the FHT calculation 218 Parameter Information dir Type int Default 0 Values forward 0 reverse 1 Demo iGraph Operators Transforms Transforms C Prototype CorOp
161. TA A AREEN VENE AEREN AA eg od Bae 22 SS CON POEPEN AEE E ENE EAEE E EAEE T THAR CTU 2 sso sc Sain ee ANE E aa 23 strLength makes EG eri AE E E EEE AE eee 24 vectorsTolmage 51 movingAverage AA ERE IEO A EA EE 25 Display E EEE E BEO A A E TERT E E EEE TUL ARTEA EEE E E ON 26 clearDisplay NOL ceccccscccsesccccesscccesscccessscccessscccsssssccssssccesssecesssssee LO a N AET randSequence Parag a cinta hie seis aaen Df CUD Lashiranteacusentamtncersdcusmnpesssnccteasusetseseunensenrssongs asses VONGONG sven cosas E isos bee assess gs eas UNATYOD vessccseccscccsscssccssesseessesscessensecssesssesenseeasens Calibration ss ies inncecnnisin ens tannins 30 TROY sis EE E E E E wotouseconeebhaattns IM COS 32 edie Fie ol th toe ba ean Pek el eek 60 PT UILAN Coys dance oeeus ces cdeats T 104 PDL OSH srscsices cde ceaatess cutie oeeesieeetesuawinc ss teoetadoub oats 62 RCOMLNES ieran i E EE beatae oad 105 OVEVIAY DIA ernn E 63 getBoundingBoX osese 106 prompt sf TUNTIME Comhann nannaa inan 65 NOURN so Rea iaaa R Rai 107 SMALIDISPLAY sssini inr 66 hough Dir sirere 108 SUTFACE veseeeeeeeee e07 NOUQNDIPRO Lo eeeeceseeeeeeseeseeeseteceeececeeseeneeeseeess 108 terrain 1 69 Linelnter sect zcceicccccsevssvieietdezdevercstividhsvstesteestheses 109 volume 70 Merge LMES haee rE E aT ied eee 109 Edis aaas wild PUL Dh ses kB ca can ce ves bee EE 110 getCircularXEdges SWA reduce lineg nonen aa A EA aed eee 110 getConto
162. Type float levelThresh Type float subpixel Type int Default 1 Values No 0 Yes 1 lineWidth Type int Default 1 smoothing Type int Default 1 Demo iGraph Operators Edges X EdgesOnLine C Prototype CorOpRtn cor_getXEdgesOnLine CorImage In CorGraphic Line CorEdgeVector Edges CorFloatVector PathValues CorFloatVector SlopeValues int edgeType float slopeThresh float levelThresh int subpixel int lineWidth int smoothing 78 Headers include WITHOME h wEdges h Libraries WITHOME lib wObj lib WITHOME lib wEdges lib traceEdges Traces edges on the grayscale images E ms Description The traceEdges operator traces edges on the image Edges represented by chains of connected edge pixels from a grayscale image The operator s output is a vector of geometric polyline objects in which each polyline represents a chain of contiguous 8 connected edge pixels in the input image The inputs of the operator is a grayscale image Parameter thresh defines minimum magnitude value for the edge to be traced minLength is the minimum length of the continuous edge segment maxMemSize is the maximum size in bytes occupied by vector of geometric polyline objects If 0 memory is unlimited Parameter Information type Type int Default 0 Values prewitt 0 sobel 1 threshold Type float minLength Type int mode Type int Default 0 Values arbitrary 0 straight 1 maxMemSize Ty
163. Type int Default 7 kernel Type Corlmage Default 1 1 1 1 1 1 1 1 1 optimization Type int Default 1 Values preciseness 0 speed 1 Demo iGraph Operators Morphology hitormiss igr C Prototype CorOpRtn cor_morphOpen CorImage In CorImage Out int type int kernelType int width int height CorIntImage kernel int optimization CorOpRtn cor_morphOpen_consume CorImage In int type int kernelType int width int height CorIntImage kernel int optimization 150 Headers include WITHOME h wMorpho h Libraries WITHOME lib wObj lib WITHOME lib wMorpho lib outline Binary outline filter Description The outline operator generates a one pixel wide outline of objects in the binary input image The outline algorithm looks at every 3x3 neighborhood and based on white black pixel arrangements determines whether a pixel is a boundary pixel or a foreground background pixel This operator could be executed in two modes The user may choose the mode by using the optimization parameter If the parameter value is speed then execution goes way faster otherwise operator delivers more precise results Parameter Information optimization Type int Default 1 Values preciseness 0 speed 1 Demo iGraph Operators Morphology hitormiss igr C Prototype CorOpRtn cor_outline CorImage In CorImage Out int optimization CorOpRtn cor_outline_consume CorImage In i
164. UPCe operator recognizes UPC E barcode See getUPCa help for details Parameter Information noise Type int Default 5 Demo iGraph Operators Matrix UPSe igr C Prototype CorOpRtn cor_getUPCe CorImage Target CorObj ROI CorObj Decoded int noise Headers include WITHOME h wMatrix h Libraries WITHOME lib wObj lib WITHOME lib wMatrix lib gradeECC200 Grades the quality of the ECC200 barcode 232 a re bi 3 Description The gradeECC200 operator grades ECC200 barcode It uses getECC200 outputs and grayscale image of the ECC200 barcode as an inputs See getECC200 help for details Explanation of the outputs and parameters could be found in the ISO standard ISO_IEC_16022 2000 Parameter dNom stands for DNOM parameter dMax stands for DMAX and parameter dMin stands for DMIN from this document cellSize is the same as in the getECC200 Parameter Information dNom Type float Default 1 dMax Type float Default 1 5 dMin Type float Default 0 5 cellSize Type int Default 10 Demo iGraph Operators Matrix UPSe igr C Prototype CorOpRtn cor_gradeECC200 CorImage Target CorStringVector Decoded CorImageVector Test CorGraphicVector Info int Contrast int PrintGrowth int AxialNonunfrmty int UnusedErrCor int Overall float dNom float dMax float dMin int cellSize Headers include WITHOME h wMatrix h Libraries WITHOME lib wObj lib WITH
165. Values 8 0 16 1 32 2 64 3 128 4 256 5 C Prototype CorOpRtn cor_ocrPerimInit CorOcrFont Font CorFontModel FontModel int samples Headers include WITHOME h wOcr h Libraries WITHOME lib wObj lib WITHOME lib wOcr lib 241 ocrPerimRun Perform optical character recognition using character perimeters EE a3 a 3 Description The ocrPerimRun operator performs optical character recognition on a binary image using a precalculated font model The operator matches specified blobs connected sets of pixels in the target image to characters in the font model Matching between the blobs and the font model characters is based on the blobs perimeters Consequently in order to perform effective matching the characters in the binary image must be separated and complete Inputs Target The target image must be a binary image Binary images can be produced by the ocrBin ocrRoiBin and ocrLocalBin operators ROI The object that specifies the blobs of interest can be a region of interest ROD or vector of ROIs in the image or a vector of Blob features The ROI or ROIs are a rectangle type Graphic object or a vector of rectangle type Graphic objects A single rectangle can be created using the createRect operator by specifying its corner points The entire image can be specified using the getRect operator A vector of ROIs can be created interactively using the getData operator If a vector of
166. Values Window 0 Weighted 1 Demo iGraph Operators Calculate Lut C Prototype CorOpRtn cor_movingAverage CorImage In CorImage Avg int number int type Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib mul Multiply two objects Description Multiply two objects See add for more information about supported types 26 Demo iGraph Operators Calculate Lut C Prototype CorOpRtn cor_mul CorObj A CorObj B CorObj product Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib not Logical NOT Description The not operator takes the logical NOT of an input integer object and sends the result to its output Demo iGraph Operators Calculate Lut C Prototype CorOpRtn cor_not int In int Out Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib randSequence Produce a vector of random integer or float numbers Description The randSequence operator produces a vector of random integer or float numbers from a uniform distribution over a specified range The number of values produced is specified by the number parameter The type of the output vector elements is specified by the type parameter and the range of numbers from which they are selected is specified by the low and high parameters The range for integers
167. When this parameter is set to Float the output vectors contain the pixel sums with the first sum corresponding the the top row or leftmost column When this parameter is set to RealPoint the output vectors contain Real points with the pixel sums in the y field and the row or column number in the x field If the useMask parameter is set to No the projections are calculated using the entire image ROI If the useMask parameter is set to Yes the projections are calculated using pixels in the intersection of the image ROI and a mask specified by the Mask parameter The mask can be specified by a Geometry or Graphic object or by a binary image The mask may also be specified by a vector of Geometry or Graphic objects binary images or Blobs In this case pixels may be counted more than once if the elements overlap in the image Parameter Information output Type int Default 0 Values Float 0 RealPoint 1 useMask Type int Default 0 Values No 0 Yes 1 Mask Type CorObj Default OBJ_B T CorGeom 4 CorFpoint 2 0 0 99 99 Demo iGraph Operators Statistics Projection Operators Statistics LocateInProj_Slope Operators Statistics LocateInProj_Centroid C Prototype CorOpRtn cor_projection_1 CorImage In CorVector Horizontal CorVector Vertical int output int useMask CorObj Mask Headers include WITHOME h wStats h Libraries WITHOME lib wObj lib WITHOME lib wStats lib smoothHist smoot
168. a labeled image in which each blob is assigned a distinct non zero value The binary and outline settings produce binary images of the filled blobs and blob outlines respectively The blobs are assigned the value of their label when this parameter is set to label value The final setting may be used when the labeled mode of the webGetBlobs or getBlobs operator is used to produce the input blobs Parameter Information yOffset Type int Default 0 labelType Type int Default 0 Values greyscale 0 binary 1 outline 2 label values 3 C Prototype CorOpRtn cor_webLabel1Blob CorImage BlobSource CorBlobVector Blobs CorImage LabeledImage int yOffset int labelType 282 Headers include WITHOME h wWeb h Libraries WITHOME lib wObj lib WITHOME lib wWeb lib webLineCounter Provide parameters and control signals for streams of images a ii p Description The webLineCounter operator provides control parameters and signals for the web operators This operator is used where each image in a stream of images represents a vertically adjacent blocks from a larger image space composed of a known number of blocks The larger image space may represent a distinct long image or may be part of an effectively continuous image that is broken up for convenience The operator produces the size of the current image in the stream the line number of the first line of the current image in the larger image space
169. accepts integer and float input images and does not accept color or complex images The output image is the same size as the Real Mag input image The output image ROI is the intersection of the ROIs of the two input images Image values outside the ROI are undefined Parameter Information input Type int Default 0 Values real imag 0 mag phase 1 Demo iGraph Operators Transforms CrossCorrelation C Prototype CorOpRtn cor_complexMerge CorImage Real_Mag CorImage Imag_Phase CorImage Complex int input Headers include WITHOME h wComplex h Libraries WITHOME lib wObj lib WITHOME lib wComplex lib complexSplit Split complex image into its real and imaginary component images Description The complexSplit operator converts a complex image into two float images The output parameter determines if the output images are the real and imaginary components of the complex image or the magnitude and phase The complexSplit operator is typically used to split the complex output of the fft operator into float images so they can be processed by other WiT operators The processed component images can be merged into a complex image using the complexMerge operator The output images are the same size and have the same ROI as the input image Image values outside the ROI are undefined Parameter Information output Type int Default 0 Values real imag 0 mag phase 1 Demo iGraph Operators Tra
170. age A mask can be used to specify which pixels are used to calculate the equalization mapping If the useMask parameter is set to No all pixels in the image ROI are used If the useMask parameter is set to Yes the pixels in the intersection of the image ROI and a mask specified by the Mask parameter are used to calculate the mapping The mask can be specified by a Geometry or Graphic object or by a binary image The mask may also be specified by a vector of Geometry or Graphic objects binary images or Blobs Once the mapping has been determined equalization is applied to the entire image ROI Parameter Information useMask Type int Default 0 Values No 0 Yes 1 mask Type CorObj Default OBJ_B T CorGeom 4 CorFpoint 2 0 0 99 99 Demo iGraph Operators Calculate HistEqualize C Prototype CorOpRtn cor_histEqualize_1 CorImage In CorImage Out int useMask CorObj mask Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib invert Invert an image Description The invert operator inverts the input image The input image must be an integer float or color image type Unsigned integer images and color images are inverted using the formula nz Z z where Z is the maximum pixel value for the input image type z is the old pixel value and nz is the new pixel value Signed integer images are inverted using the formula nz z 1 Float images are
171. age With an appropriate choice of kernel values the convld operator can perform such operations as smoothing edge or peak enhancement position shifting on a row by row basis The one dimensional convolution kernel is an integer vector specified by the kernel parameter The kernel size and weight values can be changed with the array editor by clicking on the modify button for the kernel parameter The entire vector is used as a one dimensional convolution kernel regardless of the number of rows specified in the array The convolution operator functions with odd or even kernel sizes The user may specify a factor using the scale parameter by which all values in the convolved image will be divided This allows the specification of normalized filters The operator will accept integer and floating point input images The output image is the same type as the input image except when filters with negative weights are applied to unsigned integer images 8 or 80 16 bit In this case the output image type will be a signed integer with the same number of bits as the input image type This may result in truncation of out of range values If truncation occurs relative image values can be maintained by increasing the value of the scale parameter or if the absolute image values are important the input image can be cast to a higher precision type 16 bit for 8 bit images float for 16 bit images Parameter Information kernel Type CorV
172. age is used The method parameter is used to select one of the sum of differences normalized cross correlation or binary comparison methods The binary and sum of differences methods are faster but normalized cross correlation is more robust in the presence of noise changing illumination or background and scale differences between the target and template The three methods use different approaches outlined below to generate a value for the quality of the match between the template and target at each pixel specified by the ROI input The sum of differences method generates a match value based on the sum of the absolute value of the differences between corresponding pixel values in the template and target The binary method generates a match value by counting the number of corresponding pixels that are either both zero or both non zero The sum of differences and binary count are normalized so that a value of 1 indicates a perfect match and 0 indicates no match In the normalized cross correlation method the match value is the correlation coefficient between the template and the underlying region of the target The correlation coefficient is a value between 1 and 1 where values near 1 indicate strong correlation near zero no correlation and near 1 negative correlation The correlation parameter can be used with the normalized cross correlation method to specify whether to use positive correlation negative correlation or both
173. ages and graphical objects The forward matrix output at the center port specifies the transformation that takes the reference points input at the lower input port to the test points input at the upper port The reverse matrix output at the lower port is its inverse The error value output at the upper output port gives the error in pixels between the chosen test points and the transformed reference points This value can be used to detect test point sets that do not produce a good match The getRigidBodyTransform operator is usually used to generate matrices for use with the warp operator The matrices may also be used with the warpFeatures and warpPerimeter operators Parameter Information correspondence Type int Default 1 Values 1 1 0 Best 1 minAngle Type float Default 180 maxAngle Type float Default 180 C Prototype CorOpRtn cor_getRigidBodyTransform CorVector TestPts CorVector RefPts float Error CorFloatVector Forward CorFloatVector Reverse int correspondence float minAngle float maxAngle Headers include WITHOME h wGeom h Libraries WITHOME lib wObj lib WITHOME lib wGeom lib getWarpTransform Calculate the warp conversion matrix and inverse ao Description The getWarpTransform operator creates two 3x3 matrices for forward and inverse perspective transformations that relate the input point sets The output matrices can be used to transform images and graph
174. aken to insure that the filter used is the same one that was used to generated the pyramid or reconstruction will not be exact Parameter Information filter Type int Default 0 Values None 0 2x2 1 3x3 2 5x5 3 7x7 4 C Prototype CorOpRtn cor_burtRecStep CorImage Gaussian CorImage Laplacian CorImage Out int filter Headers include WITHOME h wPyramid h Libraries WITHOME lib wObj lib WITHOME lib wPyramid lib burtStep Generate a new level in a Burt pyramid a M tik Description The burtStep operator performs one iteration in the generation of a Burt pyramid see burt The image output on the top port is equivalent to applying the reduce operator to the input image The image output on the middle port is the difference between the input image and the expanded version of the reduced version of that image The bottom output port is a flag which is set to 1 if this last iteration is the final step in a FSD pyramid generation The flag is 0 otherwise The final iteration occurs when the output image size is less than minimum size in at least one dimension The filter parameter specifies the type of gaussian filter used during image reduction and expansion The filters correspond to those provide by the gauss operator within the filter library The minsize parameter denotes the minimum size of image that will be accepted as input to the operator Parameter Information filter Type in
175. alcThresh Type int Default 1 Values No 0 Yes 1 thresholds Type CorVector threshFactor Type float Default 0 7 C Prototype CorOpRtn cor_ocrRoiBin CorImage In CorObj ROI CorImage Out CorFloatVector Thresholds int background int calcThresh CorFloatVector thresholds float threshFactor Headers include WITHOME h wOcr h Libraries WITHOME lib wObj lib WITHOME lib wOcr lib ocrRun Recognize characters in an image wait a ky ree Description See SmartOCR manual for details C Prototype CorOpRtn cor_ocrRun CorImage Image CorOcrinfo Info CorString Result CorStringVector Characters CorFloatVector Scores CorFpointVector Locations CorGraphicVector Gfx Headers include WITHOME h wOcr h Libraries WITHOME lib wObj lib WITHOME lib wOcr lib 247 ocrSelectFont Select a subset of a font s characters he P E Description The ocrSelectFont operator selects a subset of characters from from a OcrFont object based on their labels Characters in the input font with labels that match one of the strings in the labels parameter are collected into an OcrFont that is output at the Selected output All characters that do not match one of these strings collected into an OcrFont that is output at the Remaining output Label matching is case sensitive The OcrFont models are usually produced by the ocrMakeFont operator and are used by th
176. alling 1 unknown 2 subpixel Type int Default 0 Values No 0 Yes 1 numLines Type int Default 3 lineWidth Type int Default 1 smoothing Type int Default 1 Demo iGraph Operators Edges XEdgesInPoly C Prototype CorOpRtn cor_getXEdgesInPoly CorImage In CorObj Polygon CorEdgeVector PtEdges CorEdge LineEdge int NFound float slopeThresh float levelThresh int edgeType int subpixel int numLines int lineWidth int smoothing Headers include WITHOME h wEdges h Libraries WITHOME lib wObj lib WITHOME lib wEdges lib getXEdgesOnLine Find edges crossing a specified line in the image Description The getXEdgesOnLine operator finds step edges crossing a given line in an image The inputs to the operator are an integer or float image at the upper input port and line type graphic object at the Line input port The operator produces a vector of Edge objects that meet the criteria specified by the operator s parameter values Edges are located at local slope maxima or minima with an absolute value greater than the slopeThresh parameter value and with an associated pixel value level change that is greater than the levelThresh parameter value Each Edge object s slope field contains the maximum positive slope value for rising edges and the minimum negative slope value for falling edges The level contains the level change associated with the slope measured from the nearest
177. alue of 255 for bits Parameter Information port Type uint Default 0 bits Type uint Default 255 direction Type int Default 0 Values Read 0 Write 1 C Prototype CorOpRtn acqIoSetup uint port uint bits int direction Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqloWrite Write to external I O on frame grabber Description The acqloWrite operator supports frame grabbers that have digital I O output capability This operator 295 is used to set output lines on a digital output port given the port and bits pattern The mask parameter controls which lines will be set Those lines whose corresponding bit in the mask are 1 will be written to Lines whose corresponding bit in the mask are 0 will not be affected by the execution of the operator The acqloWrite operator has two modes controlled by the mode parameter In immediate mode the output lines selected by mask are immediately set to the values given by bits and the operator returns The lines remain in that state until subsequent execution of another acqloWrite or acqIoRelease operator In delayed mode the bahavior is more complex Delayed mode operates in conjuntion with a hardware event counter available on some frame grabber boards When the operator executes the output lines selected by the mode parameter are immediately set to the complement of the value given by the b
178. ameter is set to Yes the histograms are calculated using pixels in the intersection of the image ROI and a mask specified by the Mask parameter The mask can be specified by a Geometry or Graphic object or a binary image The mask may also be specified by a vector of Geometry or Graphic objects binary images or Blobs In this case pixels may be counted more than once if the elements overlap in the image Parameter Information type Type int Default 0 Values normal 0 cumulative 1 nbins Type int Default 256 lo Type float Default 0 hi Type float Default 255 useMask Type int Default 0 Values No 0 Yes 1 Mask Type CorObj Default OBJ_B T CorGeom 4 CorFpoint 2 0 0 99 99 Demo iGraph Operators Statistics Histogram Operators Calculate HistEqualize Operators Calculate AddNoise C Prototype CorOpRtn cor_histogram_1 CorImage In CorFpointVector Hist int type int nbins float lo float hi int useMask CorObj Mask Headers include WITHOME h wStats h Libraries WITHOME lib wObj lib WITHOME lib wStats lib projection Compute row and column sums alra i pe as Description The projection operator produces row and column sums for the input image The vector produced at the Horizontal output port contains the sums of the rows the vector produced at the Vertical port contains the sums of the columns The output type can be selected with the output parameter
179. ameter specifies the method used to search the local neighbourhoods When the parameter is set to Fast a binary hill climbing method is used The hill climbing method can get stuck in local a maximum in particular when the target and pattern model images contain fine regular patterns To prevent this occuring the parameter can be set to All in which case an exhaustive search of the neighbourhood is performed The exhaustive search is slower for large search ranges The MultipleMatches Sub Panel The MultipleMatches sub panel includes parameters that control the output when multiple matches are required max Matches The maxMatches parameter sets the maximum number of matches that will be produced by the operator The operator may produce fewer matches relThresh The relThresh parameter is used to set a relative rejection threshold for the matches The relative rejection threshold is determined by multiplying the parameter value by the match score of the best match found This value is then applied like the normal rejection threshold selectMode The selectMode parameter controls how the operator treats multiple matches that may correspond to the same target When this parameter is set to Location any matches found within the range specified by the xVicinity and yVicinity parameters are considered duplicates When it is set to Pattern any matches that correspond to the same element in the pattern model vector are considered d
180. an Description The acqColorkey operator is available for frame grabbers that support live video display and color keying Any pixel set to the color key value becomes a live video pixel otherwise the actual pixel value is displayed This is how graphics text or images can be drawn on a live video window to 291 report results with the live video data shown underneath Depending on the VGA color depth the colorkey parameters are set differently If you are using 256 colors then only the red parameter is used to select an index into the colormap which will be used for the color key value If you are using more than 256 colors then the all parameters are used to set an RGB value to act as the colorkey value Parameter Information red Type int Default 100 green Type int Default 255 blue Type int Default 50 C Prototype CorOpRtn acqColorkey int red int green int blue Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqCounterRead Read current value of acquisition event counter Description The acqCounterRead operator reads the value of the hardware acquisition event counter if available The acqCounterReset operator must execute at least once before acqCounterRead executes for the first time 292 C Prototype CorOpRtn acqCounterRead int Count Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib
181. an filtering on greyscale images The operator sets each pixel in the output image to the median value determined over all selected pixels in the neighbourhood of that pixel The pixels to be used are selected using a structuring element defined by the kernel parameter For more on the use of the kernel consult the documentation or help for the berode and bdilate operators The median operator is useful for removing isolated lines of pixels while preserving spatial resolutions Its performance is poor when the number of noise pixels in the neighborhood is greater than half the number of pixels in the neighborhood Median filtering performs very well on images containing binary noise but performs poorly when the noise is Gaussian For example the kernel ror Oro eror is very good at removing vertical and horizontal binary noise lines This operator could be executed in two modes The user may choose the mode by using the optimization parameter If the parameter value is speed then execution goes way faster otherwise operator delivers more precise results Parameter Information kernel Type Corlmage Default 1 1 1 1 1 1 1 1 1 optimization Type int Default 1 Values preciseness 0 speed 1 Demo iGraph Operators Morphology hitormiss igr C Prototype CorOpRtn cor_median CorImage In CorImage Out CorIntImage kernel int optimization CorOpRtn cor_median_consume CorImage In CoriIntImage k
182. an h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib 167 renumber Sequentially number named fields Description The renumber operator sets the value of the specified field in each element vector of complex objects to a sequence number The field is specified by the field parameter and must be one of the numeric basic types or a string The value assigned to the field in the first vector element is specified in the from parameter and is incremented by one for each subsequent element Parameter Information field Type CorString Default serial from Type int Default 1 C Prototype CorOpRtn cor_renumber CorVector In CorVector Out CorString field int from Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib selectElem Select an element based on the value of named field 168 Description The selectElem operator allows the user to select the maximum or minimum element or the element with value nearest to a specified value from a vector of simple numeric objects or complex objects Selection from a vector of complex objects is based on the value of a named field within the object specified by the field parameter The field must be one of the numeric or character basic types If the elements of the input vector are simple numeric types the the field parameter must be left empty The mode of operation is selected using the
183. an input vector in which each pattern model vector represents a different scaling of a pattern image and each pattern model within a vector represents a different rotation of the scaled image The patRotateScale operator and the patInit2D1 operator can be used together to produce a vector of pattern model vectors in this format from a single image When applicable the pattern models should be more closely related within each pattern model vector than they are across the vector of vectors For example each vector in the top level vector might represent a different character while the pattern models within a vector represent various forms or degraded versions of that vector s character The patterns may be broken into equal sized rectangular tiles and or may incorporate an arbitrary binary mask as described in the patInit2D1 operator documentation Locations The match locations are produced as a vector of Real points giving the x and y coordinates of the location in the matching region in the target image corresponding to the center pixel in the pattern image For even dimension width or height pattern images the center pixel location is rounded up to the next integer value Only the best match at each location is reported The matches are sorted in order of descending score Scores The match scores corresponding to the locations in the Locations output vector are produced as a vector of float values Indexes1 and Indexes2
184. an the 0 5 pixels that might be expected due to the limitations of the rotation algorithm used This may result in noticeable artifacts particularly in rotation of thin lines The background regions in the output image the regions that correspond to areas outside the input image must be filled with an appropriate value When the background parameter is set to auto the background value is computed as the average of the four corner pixels in the original image When it is 130 set to user the background value is set explicitly using the bgValue parameter Parameter Information angle Type float Default 0 sizeOp Type int Default 0 Values expand 0 clip 1 rotateAbout Type int Default 0 Values center 0 user 1 rotatePt Type CorPoint Default 0 0 antiAlias Type int Default 1 Values no 0 yes 1 background Type int Default 0 Values auto 0 user 1 bgValue Type float Default 0 Demo iGraph Operators Fit polar igr C Prototype CorOpRtn cor_rotate CorObj In CorObj Out float angle int sizeOp int rotateAbout CorPoint rotatePt int antiAlias int background float bgValue Headers include WITHOME h wGeom h Libraries WITHOME lib wObj lib WITHOME lib wGeom lib scroll Scroll an object Description The scroll operator shifts the object along the y axis by the given amount The amount paramater controls the number of pixels the object is shifted
185. and Y variables respectively For example the expression X gt 5 A gt gt K A will produce an output image where every pixel was right shifted by K bits unless the pixel fell within the first five columns of the image For more on the expression syntax and use see calc This operator is faster than unaryOp but does not allow the specification of the output image type 21 Parameter Information unaryOp Type int Default 0 Values 0 1 2 3 OR 4 AND 5 gt gt 6 lt lt 7 expr 8 expression Type CorString Default A Demo iGraph Operators Calculate CollectAverage C Prototype CorOpRtn cor_fastUnaryOp CorImage A CorObj Constant CorImage Out int unaryOp CorString expression Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib histEqualize Histogram equalization Description The histEqualize operator may be used to re map the frequency distribution of an image so that the image histogram is approximately flat This has the effect of spreading the range around values found in many pixels and compressing the range around values found in few pixels This often has the effect of improving perceived image contrast around the dominant values in the image It should be noted that precision of the relative image values cannot be improved by this method 22 The output image produced is the same type as the input im
186. and patRange The patCenter and patRange parameters specify elements selected from the pattern vector for the search when applyPatterns is set to Range The patCenter parameter is a vector of integers that must be the same length as the input point vector The elements of the parameter vector are indexes into the input pattern model vector that indicate the center of the range of patterns for the corresponding input point A contiguous group of 2 patRange 1 elements of the pattern vector centered on the patCenter element are used in the search Parameter Information scale range Type int Default 2 Values 1 0 2 1 4 2 8 3 16 4 32 5 Type int Default 3 Values 2 0 4 1 8 2 16 3 32 4 rejectThresh Type float Default 0 acceptThresh Type float Default 1 minContrast Type float Default 0 maxTilesOmitted Type int Default 0 localSearch Type int Default 0 Values Fast 0 All 1 maxMatches Type int relThresh Default 1 Type float Default 0 selectMode Type int xVicinity 262 Default 0 Values LocAndPat 0 Location 1 Pattern 2 Type int Default 0 yVicinity Type int Default 0 applyPatterns Type int Default 0 Values All 0 Sample 1 Range 2 patSampling Type int Default 1 patCenter Type CorVector patRange Type int Default 1 C Prototype CorOpRtn cor_patLocall1 CorImage Target CorFpointVector ROI CorPatMod
187. aryOp operator performs basic arithmetic and logical operations involving an image and a scale value Arbitrary expressions written in the C language syntax which may also include the pixel location can also be performed The operator supports grayscale and color input images If the input image is a grayscale image then the input constant must be a numeric value If the input constant is an integer the output image type is the same as the input image type If the input constant is a float or double the output image is a float image regardless of input image type Out of range values for integer images are clipped to the range of the image type If the input image is a color image then the input constant may be a color object of the same type as the input image or a numeric value In this case the output image will be the same type as the input image Supported operations for an image A and a scalar K are A K A K 7 A K A K logical OR amp logical AND gt gt shift right lt lt shift left expression utilize expression to determine result The indicated operation is performed on each pixel in A The expression field allows the user to specify a C style mathematical equation which will determine the output image The image can be specified in this equation using the variable A while the constant is specified using K The result for each pixel can be affected by that pixel s particular column and row using the X
188. as the y coordinate It is also possible to plot a multi line graph from a compound vector vector of vectors or from multiple individual vectors described below When input data arrives a graph frame pops up displaying the graph as specified by the operator s parameters A properties panel can be invoked from the pop up menu inside the graph region and can be used to change the method of displaying the graph data Pressing the left mouse button while the cursor is on the graph gives the exact cursor position in graph coordinates The graph can be scaled and panned in the same manner as an igraph Graph is a powerful and often used tool See chaper Operators Libraries in the User s Manual for details 57 Parameter Information giTitle name type cumulative allowDuplicateNames showDataPts showLegend legendPos xaxis axisXScale xaxisRange xShowTicks yaxis axisYScale yaxisRange yShowTicks Placement 58 Type CorString Default Graph Type CorString Default Curve Type int Default 0 Values Line 0 Bar Chart 1 Histogram 2 Scatter Plot 3 Type int Default 0 Values No 0 Yes 1 Type int Default 1 Values No 0 Yes 1 Type int Default 0 Values No 0 Yes 1 Type int Default 0 Values No 0 Yes 1 Type CorPoint Default 0 0 Type CorString Default x axis Type int Default 0 Values Auto 0 Manual 1 Type Co
189. at Default 1 getData2 Interactive specification of graphical objects ae th F xs Description The getData2 operator behaves like the getData operator except that it accepts a second input which contains an initial set of graphic objects vector of Graphic This input can be a sync token also which allows getData2 to fire with a blank set of graphics data Parameter Information name Type CorString Default getData2 startMode Type int Default 0 Values Data entry 0 Select 1 Always selected 2 Select first 3 type Type int Default 0 Values point 0 line 1 polyline 2 polygon 3 rectangle 4 circle 5 text 6 Placement Type int Default 0 Values bitwise OR Fix position 1 X Type int Default 0 Y Type int Default 0 XScale Type float Default 1 YScale Type float Default 1 graph Create line bar histogram or scatter plots Description The graph operator accepts either a single vector or compound vector vector of vectors of values for plotting single or multi line graphs The operator can generate line graphs bar charts histograms and scatter plots HT ait 3 The graph operator accepts vectors of integer points or float points or vectors of integer or float scalar numbers Point input data is plotted using the x y coordinates from the data Scalar data is plotted using the vector index starting from 0 as the x coordinate and the element value
190. ate Blob features of warped objects Description The getWarpedFeatures operator computes Blob feature vectors in a perspective warped image space for each blob in the blob vector input The blob vector is produced by applying the getBlobFeatures operator to the original unwarped binary image The feature values are calculated in a warped image space which is specified by a transformation matrix input atthe lower input port of the operator The transformation matrix is typically produced by the getWarpTransform operator The Blob features produced by applying the getWarpedFeatures operator to the blobs generated from the unwarped image are approximately equal to those produced by first warping the image then extracting the blobs and applying the getBlobFeatures operator The object s perimeter is not calculated The getWarpedPerimeter operator can be applied to the unwarped Blob feature vector and the unwarped image to produced the warped perimeter points The fields of the Blob feature object are described in the getBlobFeatures operator documentation Demo iGraph Operators Blobs WarpFeatures C Prototype CorOpRtn cor_getWarpedFeatures CorBlobVector Blobs CorFloatVector Transform CorFeatureVector Warped Headers include WITHOME h wBlob h Libraries WITHOME lib wObj lib WITHOME lib wBlob lib getWarpedPerimeter Find warped perimeter given a Blob feature m agis Description The getWarpedPer
191. ath is rasterized or sampled at equally spaced points The rasterize option returns pixel locations corresponding to a single pixel width curve in the image as would be used in rendering the graphic object Each location correspond exactly to a pixel and no pixel appears more than once except possibly in self crossing polylines or polygons It should be noted that the sampling distance is dependent on the orientation of the curve For example a horizontal line will have approximately 1 4 times as many sample points as a line of the same length at a 45 degree angle Selecting the sample option for the operation produces pixel locations corresponding to equally spaced samples along the graphical object The distance between samples is specified using the sampleSpacing parameter and is given in pixel units For lines polylines and polygons equally spaced sample points are determined starting from the first end point of the graphical object The last end point and intermediate segment end points will not necessarily be sample points The path for a circle starts at the point x_c r y_c where x_c y_c is the center of the circle and r is the radius and proceeds clockwise around the circumference of the circle The equally spaced sample locations do not necessarily fall exactly on pixel centers The method of assigning a value from the image to a pixel location is controlled by the sampling parameter If the nearest option is se
192. ation dir Type int Default 0 Values E W 0 NE SW 1 N S 2 NW SE 3 optimization Type int Default 1 Values preciseness 0 speed 1 Demo iGraph Operators Filter Entropy Operators Filter V ariance C Prototype CorOpRtn cor_line CorImage In CorImage Out int dir int optimization CorOpRtn cor_line_consume CorImage In int dir int optimization Headers include WITHOME h wFilter h Libraries WITHOME lib wObj lib WITHOME lib wFilter lib 88 lopass1d One dimensional image smoothing TE 5 E a Description The lopassild operator applies one dimensional low pass filtering smoothing to each row of the input image The effect of the smoothing operation is to blur the image in the x direction reducing the sharpness of vertical edges and fine detail in the image in the x direction and reducing the effects of high frequency noise One dimensional low pass filtering is faster than two dimensional low pass filtering Also since the filtering is only performed along the X axis original image information along the Y axis is preserved The smoothing operation is performed with one dimensional a moving average kernel of a size in pixels specified by the width parameter The effect of lopassid is equivalent to that of the lopass2d with a height of 1 The operator accepts integer and float type input images The output image is the same type as the input image Parameter I
193. ation used to calculate the offsets The limit on the offset for a threshold can be determined by multiplying this value by the loOffset or hiOffset parameter value See the thresh operator for a description of the invert levels and result parameters and the output images that result from various combinations of these parameters MMX acceleration is available for 8 bit unsigned input images in the mean min mode when the loOffset is 0 5 0 25 0 125 or 0 0625 invert is true levels is set to single result is binary the size of the kernel width times height is a power of two not greater than 256 and both width and height are greater than one Parameter Information width Type int Default 25 height Type int Default 25 offsetMethod Type int Default 0 Values absolute 0 mean min 1 max mean 2 standard deviation 3 loOffset Type float Default 0 hiOffset Type float Default 0 minOffsetBase Type float Default 0 invert Type int Default 0 Values no 0 yes 1 levels Type int Default 1 Values double 0 single 1 result Type int Default 1 Values semi threshold 0 binary 1 Demo iGraph Operators Segmentation LocalAdaptThresh 193 C Prototype CorOpRtn cor_localAdaptThresh CorImage In CorImage Out int width int height int offsetMethod float loOffset float hiOffset float minOffsetBase int invert int levels int result Headers include WITHOME h wSegment
194. ator takes a WiT image converts it to a file in a standard image format specified by the filename parameter The object directory is used as the directory to write the file unless an absolute path name is given The type parameter specifies the image format Currently the following formats are supported Graphics Interchange Format 8 rx frci no __ supported pcx Pe PaintBrush 8 PC PaintBrush 8 bi Portable Bitmap Format PBM sau kai kaasa xs R a Format TIFF o Tag Image File Format Image File Format Microsoft Windows Device ey Bitmap Format PEG PEG Joint Joint Photographers Expert Group Expert Group WiT WiT object format images only Raw raw data of known size and type possibly with a header of known length JPEG2000 JP2 JP2 JPEG 2000 20 00 2000 JPEG2000 JPEG 2000 CodeStream CodeStream The dataStorage and bitmap options are only used for PBM formats and will be explained below The quality parameter is used only for the JPEG format For PCX formats writeImage does not support 16 bit images The operator scales 16 bit images to 8 bit images before writing to disk If a 24 bit color image has more that 256 colors writeImage quantizes the colors in order to fit into the colormap Color deterioration and longer processing time should be expected However the trade off is a much smaller file size For PBM format writeImage converts a WiT image to bit mapped PBM greyscale PGM
195. ault 0 y Type int Default 0 w Type int Default 256 h Type int Default 256 C Prototype CorOpRtn acqSize int x int y int w int h Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqSnapCount Set frame grabber s acquisition snap count Description The acqSnapCount operator controls the number of frames acquired by a frame grabber for each frame delivered by the frame grabber to WiT This is only available for certain frame grabbers The value should not normally be modified at run time It may be necessary to change the snap count after loading certain sophisticated custom pixel processor designs Parameter Information snapCount Type uint Default 1 C Prototype CorOpRtn acqSnapCount uint snapCount Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqSource Switch input source on frame grabber 140 E Description The acqSource operator is used to switch the video input source from one camera to another for the current frame grabber The source parameter specifies the desired input channel Input channels begin at zero Parameter Information source Type int Default 0 C Prototype CorOpRtn acqSource int source Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqStart Start ping pong acquis
196. ault 2 threshold Type float Default 0 ignoreNoMatch Type int Default 0 Values No 0 Yes 1 noMatchLabel Type CorString Default selection Type int Default 0 Values bitwise OR Max size 1 Min size 2 minWidth Type int Default 0 maxWidth Type int Default 0 minHeight Type int Default 0 maxHeight Type int Default 0 minPixels Type int Default 4 maxPixels Type int Default 0 C Prototype CorOpRtn cor_ocrPerimRun CorImage Target CorObj ROI CorFontModel FontModel CorVector Labels CorVector Scores CorVector Locations int rotation float sensitivity float minAngle float maxAngle float featuresUsed int scaleInvariant float minScale float maxScale float threshold int ignoreNoMatch CorString noMatchLabel int selection int minWidth int maxWidth int minHeight int maxHeight int minPixels int maxPixels Headers include WITHOME h wOcr h Libraries WITHOME lib wObj lib WITHOME lib wOcr lib ocrPlotFont Create graphic representation of the font from the font model 5 Description The ocrPlotFont operator creates a graphic representation of the sampled perimeters used to extract features from the font model Inputs FontModel The input to the operator is a font model created by the ocrInitialize operator Outputs Perimeters The output is a vector of polygon Graphic objects in which each element represents one character
197. bHeight int reset int resetBlobId 279 Headers zero connected pixels are considered to belong to the same connected blob In a include WITHOME h wWeb h labeled image equal valued non zero connected pixels belong to the same blob The input image is usually produced from Libraries an original image in a stream of images processed by the web operators or from WITHOME lib wObj lib the output of the webWatershed WITHOME lib wWeb lib operator Outputs 5 Out webGetPerimeter The output Feature vector is a duplicate of the input Feature vector with the Feature Find perimeter given a feature in a waterfall image perimeter data set added to each element The perimeter data set values are in the i frame of reference of the original feature gi a set The perimeter fields are defined as k SIN follows Description Total length of The webGetPerimeter operator determines the ante Gi of points perimeter related data corresponding to an input around onec Feature vector and an input image from a frame of maxDeltaX Max distance along x reference shifted in y relative to the Features The axis from x centroid perimeter is determined for each element in the input vector by tracing equivalent valued boundary minDeltaX Min distance along ae pixels in the input image axis from x centroid maxDeltaY Max distance along y This operator is typically used to calculate axis from y centroid erimeters for Feature vectors genera
198. baudRate Type int Default 9600 parity Type int Default 0 Values none 0 odd 1 even 2 mark 3 space 4 stopBits Type int Default 0 Values 1 0 1 5 1 2 2 dataBits Type int Default 3 Values 5 0 6 1 7 2 8 3 setTimeOuts Type int Default 0 Values No 0 Yes 1 rdTimeOutMult Type int Default 0 rdTimeOutConst Type int Default 0 rdIntervalTimeOut Type int Default 0 wrTimeOutMult Type int Default 0 wrTimeOutConst Type int Default 0 C Prototype CorOpRtn cor_openSerial 289 CorObj In CorObj Out int port int configurePort int baudRate int parity int stopBits int dataBits int setTimeOuts int rdTimeOutMult int rdTimeOutConst int rdIntervalTimeOut int wrTimeOutMult int wrTimeOutConst Headers include WITHOME h wSerial h Libraries WITHOME lib wObj lib WITHOME lib wSerial lib closeSerial Close a serial port a nt Description The closeSerial operator closes a serial port that has been previously opened for reading and writing The input object is passed through to the output unchanged The port parameter is used to specify the input port to be closed The port parameter value 0 corresponds to the COM1 port parameter value 1 corresponds to COM2 and so on If the specified port is not open the operator does nothing Ports are not automatically closed when an igraph stops running Parameter Information port Type int De
199. be the same size and type Outputs Out The output images are the same width and type as the input images The height of the output image is the height of the input images plus two times the number of specified number of rows The output image has the appropriate number of rows from the current image appended to the bottom of the image In the usual case where the height of the input image is greater than the number of rows added this results in a delay of one image as the stream passes through this operator The first output image will only contain the top rows of the first input image in the stream and most of the last image will not be output Parameters 286 rows The rows parameter specifies the number of rows by which an image is extended at the top and bottom Parameter Information rows Type int Default 2 C Prototype CorOpRtn cor_webRowExtend CorImage In CorImage Out int rows Headers include WITHOME h wWeb h Libraries WITHOME lib wObj lib WITHOME lib wWeb lib webRowRemove Remove rows from the top and bottom of images yo ed Description The webRowRemove operator strips a specified number of rows from the top and bottom of the input image This operator is usually used following the webRowExtend operator to return the extended images to their original size Inputs In The input images may be any type The image height must be greater than twice the number of rows s
200. braries WITHOME lib wObj lib WITHOME lib wMatrix lib getECC200 2 Find and decode ECC 200 2D barcodes Description The getKCC200_1 operator recognises ECC200 barcode See getECC200 help for detales Differ from getECC200 this operator has two images as an inputs One is a binary image for position finding and another is a grayscale image for the rest of the recognition tasks Output of this operator is a vector of vectors of bytes rather then the vector of CorStrings 227 Parameter Information cellSize Type int Default 10 Demo iGraph Operators Matrix ECC200 igr C Prototype CorOpRtn cor_getECC200_2 CorImage Target CorImage Binary CorVector DecodedBin CorImageVector Test CorGraphicVector Info int cellSize Headers include WITHOME h wMatrix h Libraries WITHOME lib wObj lib WITHOME lib wMatrix lib getiTF Finds and decodes interleaved 2 of 5 bar code o a MIEG Description The getITF operator recognizes Interleaved 2 from 5 ITF barcode See getCode39 help for more details Parameter checkSum defines whether checksum should be used Parameter Information noise Type int Default 5 checkSum Type int Default 1 Values No 0 Yes 1 228 Demo iGraph Operators Matrix ITF igr C Prototype CorOpRtn cor_getITF CorImage Target CorObj ROI CorObj Decoded int noise int checkSum Headers include WITHOME h wMatrix h Libraries
201. c data to lines or circles and operators that find centroids bounding boxes and convex hulls of image or geometric data cHough Circular Hough transform ETI lt s Description The cHough operator computes the forward or reverse circular Hough transform of the input The direction parameter specifies whether the forward or reverse transform is performed The forward circular Hough transform of a binary image unsigned 8 or 16 bit is a three dimensional function of x y and r radius giving the fraction of pixels on a circle of radius r centered at x y that are set This transform can be used to locate circles in images The forward transform generates a vector of unsigned 8 bit images where each image in the vector corresponds to circles of one radius A pixel value in an output image represents the fraction of set pixels in the input image on a circle centered at the corresponding pixel with a radius equal to the radius value associated with the image The pixel values are normalized to a value of 255 for a complete circle The radius values corresponding to each image range from startRadius to endRadius in one pixel steps For example pixel x y of image i in the output vector gives the fraction of pixels on a circle in the input image centered at x y with radius i startRadius that are turned on scaled to the range 0 to 255 The reverse transform converts points in the input image to circles in the output image I
202. center mouse button Once the object has been drawn another object of the same type can be entered or you can change the object type using the Graphics Editor Completed graphical objects are displayed overlayed on the image The completed objects can be adjusted moved or deleted by clicking on the select button the arrow icon at the left of the lower row of the editor panel An object is selected by clicking on the object causing the object to turn red A selected object can be moved by dragging it with the mouse or can be deleted by clicking on the delete button the scissors icon at the left of the upper row of the editor panel After clicking the select button objects can be resized by pressing the left mouse button near the part of the object to be modified and dragging it to the desired position the object must not be selected before you begin this operation The cursor will turn into a screwdriver icon in this mode The editor panel also allows the specification of outline or fill mode for the objects and the color associated with each as well as control of the fonts for text Parameter Information name Type CorString Default getData type Type int Default 0 Values point 0 line 1 polyline 2 polygon 3 rectangle 4 circle 5 text 6 Placement Type int Default 0 Values bitwise OR Fix position 1 X Type int Default 0 Y Type int Default 0 XScale Type float Default 1 YScale Type flo
203. ch fall outside the image type s range are clipped to the range The random number sequence used to generate the noise can be controlled using the useSeed seed and reset parameters A random number seed is generated the first time an operator executes or when the reset parameter is set to Yes starting a new random number sequence When the reset parameter is set to No the random number sequence is a continuation of the sequence used the last time the operator executed When the useSeed parameter is set to No the initial random number seed is generated from a value derived from the system clock so a different random number sequence is produced each time the graph is run When the useSeed parameter is set to Yes the value specified in the seed parameter is used to initialize the random number sequence producing repeatable noise patterns in the image The seed value is a three element vector of short integers The operator s NextSeed output value is the seed value that will continue the random sequence used by the operator Parameter Information affectedArea Type int Default 50 Values 0 100 noiseType Type int Default 0 Values random 0 spike 1 gaussian 2 noiseAmount Type float Default 100 useSeed Type int Default 0 Values No 0 Yes 1 seed Type CorVector Default 1 0 0 reset Type int Default 0 Values No 0 Yes 1 Demo iGraph Operators Calculate AddNoise Operators Calculate Col
204. cleCenters C Prototype CorOpRtn cor_getPeaks CorImage i0 CorVector Peaks float threshold float relThreshold int maxPeaks int type int subpixel Headers include WITHOME h wFind h Libraries WITHOME lib wObj lib WITHOME lib wFind lib nxcorr Calculate normalized cross correlation of target and template m E mi ia Description The nxcorr operator produces the normalized cross correlation image from a target image and a smaller template image The template image used is the ROI of the Template input image A pixel value in the output image represent the normalized cross correlation value calculated with the upper left corner of the template image at the corresponding pixel in the input target image Values are only calculated when the template is entirely inside the target image ROI so the template can be no larger than the target ROI Demo iGraph Operators Find NXCorr C Prototype CorOpRtn cor_nxcorr CorImage Target CorImage Template CorImage Nxcorr Headers include WITHOME h wFind h Libraries WITHOME lib wObj lib WITHOME lib wFind lib sod Calculate sum of differences between target and template Description The sod operator produces the sum of differences image from a target image and a smaller template image The template image used is the ROI of the Template input image A pixel value in the output A pixel value in the output image represe
205. clude WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqLineTrigger Control a frame grabber s line trigger mode Description The acqLineTrigger operator is used to control the line trigger method for a linescan camera attached to the currently selected frame grabber When the source is set to Internal the frame grabber board will generate line sync signals for the camera at a fixed line rate If the source is External or Encoder the board will generate line sync signals for the camera based on an external line trigger signal supplied to the board If the source is None the board will send no line sync signals to the camera If the source is Encoder then the dropCount parameter determines how many encoder pulses are ignored before the board sends a line sync signal to the camera For example if dropCount 7 then the camera will acquire a single line for every 8th encoder pulse Parameter Information source Type int Default 1 Values None 0 Internal 1 External 2 Encoder 3 dropCount Type int Default 0 Demo iGraph Operators Video Acquisition trigger C Prototype CorOpRtn acqLineTrigger int source int dropCount Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqLive Turn on off live video display ache Description The acqLive operator is supported for frame grabbers that can display live v
206. cor_directory CorStringVector Filename CorFile filename Headers include WITHOME h wSystem h Libraries WITHOME lib wObj lib WITHOME lib wSystem lib exit Exit WiT Description The exit operator causes WiT to exit when it receives any object at its input If the confirm parameter is set to no WiT will terminate immediately If the parameter is set to yes the user will be prompted before termination Parameter Information confirm Type int Default 0 Values Yes 0 No 1 206 fileStatus Check the status of a file A Description The fileStatus operator reports the existence size or time of last modification of a file The name of the file to be tested is specified using the filename parameter The request parameter determines which of the three status values is reported If it is set to Exists the operator will output a one if the file exists or a zero if it does not If the request parameter is set to Size the operator will output the size of the file in bytes If the named file does not exist the operator will output 1 If the request parameter is set to Time the operator produces the time at which the specified file was last modified The format of the output is controlled by the format parameter The Integer format output is the number of seconds since a standard reference date The String format output is the date and local time in the format Wed Nov 10 14 27 42 1999 I
207. core of the best candidate match found This value is then applied like the normal rejection threshold The relative threshold is used to restrict the candidates passed on to the second to those near the best applyPatterns The applyPatterns parameter is only meaningful when more than one pattern model is applied When the parameter is set to All all patterns in the input are compared to the target image in both phases of the search When the parameter is set to Sample the input vector of pattern vectors is sampled in both dimensions as specified by the patSampling1 and patSampling2 parameters for the candidate search The second phase of the search will then only use pattern vectors or patterns within the corresponding sampling distance of the vector or pattern that produced the candidate match patSampling1 The patSampling1 parameter specifies the sample spacing in the vector of pattern vectors for the candidate search when applyPatterns is set to Sample patSampling2 The patSampling2 parameter specifies the sample spacing within each vector of pattern vectors for the candidate search when applyPatterns is set to Sample output The Candidates setting of of the output parameter allows the user to output the locations match scores and pattern indexes of the candidates rather than those of the refined matches normally produced The Candidates setting is usually used during development to help determine the best settings for the
208. correspond to non zero pixels in the ROI image are used The ROI used for the line fitting is the intersection of the input ROI and the image ROI The graphic objects may be generated by the getData and createRect operators or may be constructed using the createGraphics and putField operators Binary images are often produced by one of the thresholding operators thresh iThresh or localThresh Demo iGraph Operators Geometry convexHullL igr C Prototype CorOpRtn cor_gFitLine CorImage In CorObj ROI CorLine Line Headers include WITHOME h wFit h Libraries WITHOME lib wObj lib WITHOME lib wFit lib geom2lines Convert a vector of geometric objects to a line vector Description The geom2lines operator converts a vector of line rectangle polyline or polygon Geom objects into a vector of line segments Input vector elements of other types are ignored The line segments are output as a vector of Line data objects For polyline or polygon elements of the input vector line segments are produced by determining if any vertices of the polyline are more than a specified deviation away from the line connecting its two endpoints If not the polyline endpoints are line segment endpoints Otherwise the polyline is split at the vertex farthest from the line and the algorithm is applied recursively to the two resulting polylines The deviation is specified in pixels The Line objects produced by the operato
209. counter value is always incremented every time counterStep is executed If hasFinal is yes the counter only counts to and including the value specified by the final parameter Values no 0 yes 1 final Type int Default 0 Type int See hasFinal Demo iGraph Operators clearObjs counterRecall Recall current value of counter mer ee all Description Recall current value of counter Parameter Information name Type CorString Default counter Demo iGraph Operators clearObjs counterReset Reset counter to initial value 114 nefas mab Description counter Reset resets a counter to its initial value Counters are referenced by the name parameter which refers to a global name See chapter Operators Special in User s Manual for details See also counterInit counterStep Parameter Information name Type CorString Default counter Demo iGraph Operators clearObjs counterStep Run counter for one step Description counterStep increments a counter by one step every time it executes By default a counter has an initial value of 0 and a step of 1 with no final value Use counterInit to create or re initialize a counter with non default settings The first time counterStep is executed the count output is the initial value of the counter If the step size is negative the counter counts down If the counter has a final value then the value remains at the final valu
210. cription The bmedian operator performs median filtering on binary images The operator sets a pixel in the output image true if the majority of the selected neighbourhood pixels are true and false if they are not The pixels to be used can be selected using a structuring element defined by the kernel parameter This operator could be executed in two modes The user may choose the mode by using the optimization parameter If the parameter value is speed then execution goes way faster otherwise operator delivers more precise results Parameter Information kernel Type Corlmage Default 1 1 1 1 1 1 1 1 1 optimization Type int Default 1 Values preciseness 0 speed 1 C Prototype CorOpRtn cor_bmedian CorImage In CorImage Out CorIntImage kernel int optimization CorOpRtn cor_bmedian_consume CorImage In CorIntImage kernel int optimization Headers include WITHOME h wMorpho h Libraries WITHOME lib wObj lib WITHOME lib wMorpho lib dilate Morphological dilation Description The dilate operator performs binary or gray scale morphological dilation on images Binary dilation tends to add pixels to the edges of objects in the image Gray scale erosion sets each pixel to the maximum pixel value in its neighbourhood and is sometimes refered to as a maximum filter This operator could be executed in two modes The user may choose the mode by using the optimization paramete
211. cted Feeding different object types to collector will cause an error If elementTypes is different different object types can be collected Values same 0 different 1 Demo iGraph Operators clearObjs counterlnit Create or re initialize up down counter Description Normally a counter can be used without explicitly creating it Such counters count from zero with a step increment of one every time counterStep executes counterInit is used to create or re initialize a counter with non default settings so that you can specify parameters such as the starting count step size including up or down counting and whether the count should stop counting at a specified final value Counters are referenced by the name parameter which refers to a global name See chapter Operators Special in User s Manual for details See also counterReset counterStep Parameter Information name Type CorString Default counter Type CorString The global name of the counter initial Type int Default 0 Type int The initial value of the counter The first time that counterStep is executed on a counter the output count will be value of initial incre Type int Default 1 Type int The amount that the counter value is incremented every time counterStep is 113 executed If incre is negative then the counter will count down hasFinal Type int Default 0 Type choice no 0 yes 1 If hasFinal is no the
212. ctors of the corresponding ROI in the input ROI vector Scores The operator outputs a score of for each character character found that indicates the quality of the match with the model character Scores range from 0 to 1 with 1 indicating a perfect match to the model character The scores are output as a vector of float values or a vector of float vectors depending on whether the input ROI is a single ROI or a vector of ROIs The locations have a one to one correspondence with the label strings in the Labels output Locations The operator also outputs the locations of the characters found in the target image as either a Real point vector or a vector of Real point vectors depending on whether the input ROI is a single ROI or a vector of ROIs The locations have a one to one correspondence with the label strings in the Labels output Parameters maxMatches The maxMatches parameter sets the maximum number of character matches that will be produced by the operator The operator may produce fewer matches xVicinity and yVicinity These parameters specify a minimum region around a character location in which matches with lower scores will be suppressed Two matches will never be produced that are less than both xVicinity pixels apart in x and yVicinity pixels apart in y rejectThresh The rejectThresh parameter is used to set a rejection threshold for character match scores below which no characters will be considered The pa
213. culation Given a grayscale image J the following identity holds for binary erode and gray scale erode threshold gray scale erode I binary erode threshold 1 Parameter Information type Type int Default 0 Values Binary 0 Grayscale 1 kernelType Type int Default 0 Values Rectangular 0 Arbitrary 1 width Type int Default 7 height Type int Default 7 kernel Type Corlmage Default 1 1 1 1 1 1 1 1 1 optimization Type int Default 1 Values preciseness 0 speed 1 C Prototype CorOpRtn cor_erode CorImage In CorImage Out int type int kernelType int width int height CorIntImage kernel int optimization CorOpRtn cor_erode_consume CorImage In int type int kernelType int width int height CorIntImage kernel int optimization Headers include WITHOME h wMorpho h Libraries WITHOME lib wObj lib WITHOME lib wMorpho lib hitOrMiss Hit or miss filter gt j The hitOrMiss operator performs a hit or miss type binary pattern matching operation on an image If the pixels in the neighbourhood of a given pixel match the pattern of pixels specified in the kernel then the corresponding pixel in the output image is set to the image maximum otherwise it is set to zero The kernel parameter allows the user to specify the pattern Kernel elements set to 0 match 0 value pixels in the input image kernel elements set to 1 or any positi
214. cupies one page even if it may be clipped A general object may span several printed pages If text the output is straight text ASCII Images are printed as hexadecimal values fit If Yes the object is scaled to exactly fit the page determined by pagesize and orientation If No the scale specified in scale is used scale The scale in printer points 1 72 inch per pixel color If color is chosen color images are output print to Printer send to printer file save to file EPS save to file in EPS format name name of file for file or EPS print to setting More information about printing can be found in the User s Manual Parameter Information show frame Type int Default 0 Values yes 0 no 1 format Type int Default 0 Values graphic 0 text 1 fit Type int Default 0 Values no 0 yes 1 scale Type float Default 1 print to Type int Default 0 Values printer 0 file 1 name Type CorFile Default witprint putClipboard Put an image on the clipboard Description The putClipboard operator writes an eight bit unsigned or color WiT image to the system clipboard in BMP format Images of other WiT types must first be converted to the eight bit unsigned type using the castImage operator before they are saved to the clipboard C Prototype CorOpRtn cor_putClipboard CorImage In Headers include WITHOME h wSystem h Libraries WITHOME lib wObj lib WITHOM
215. d at the beginning of the output string and the append string is inserted at the end Multi column vectors of strings are no longer supported so the rewDelimit parameter now has no effect Parameter Information colDelimit Type CorString Default rowDelimit Type CorString Default prepend Type CorString 43 Default append Type CorString Default C Prototype CorOpRtn cor_formatString CorStringVector In CorString Out CorString colDelimit CorString rowDelimit CorString prepend CorString append Headers include WITHOME h wConvert h Libraries WITHOME lib wObj lib WITHOME lib wConvert lib plotEdges Convert edge vectors to graphic descriptions a i Bf r Description The plotEdges operator converts an edge or vector of edges into a vector of graphic objects that can be displayed on an image using the overlayData operator The graphic objects include a geometric object for each edge for example a point for crossing edges found using getXEdges and may also include an arrow indicating the edge direction and a label Since multiple graphic objects are used to represent each edge a single edge input results in a vector of graphic objects output The color of the graphical objects can be selected using the parameters in the Color subpanel Parameter Information r Type int 44 g Type int Default 0 Values 0 255 b Type int Default 0 Values 0
216. d be less than the height parameter value of the webWaterfall operator by at least the height of the images in the input image stream to ensure that the blobs are contained within a single waterfall image reset The reset parameter can be used to release all active blobs and reset the running line count to zero at the end of the current image When a reset function is required this parameter is usually used as an input parameter When a zero is received the blobs are built up as usual When a one is received all blobs in the current image are output even if they reach the bottom of the image and after processing for the current image is complete the starting line number is reset to zero The webLineCounter operator can be used to provide the required control signals When no reset function is required this parameter is set to No In this mode the line count will continually increase as images are received until the graph stops Blobs which reach the bottom of the last image processed will not be output by the operator resetBlobld The blob id number can be set to zero with the resetBlobId parameter This will set the next blob id to zero after processing the current image This parameter is usually set to No or used as an input parameter It may be tied to the reset parameter when both are input parameters in which case a new set of id numbers will be started every time the operator is reset 277 Parameter Information conn
217. d by the operator to see if they extend into the next image in the sequence The operator can be forced to output these blobs using its parameters The operator is approximately equivalent to applying the getBlobsROI operator to an image formed by concatenating the images in the stream into a single image The difference is that this operator provides parameters to reset the operator s state and to limit the maximum height of the blobs produced Inputs BinaryImage The BinaryImage input must be an 8 or 16 bit unsigned image The images may be interpreted as either binary images such as those produced by the thresh operator or labeled images such as those produced by the kMeans operator All images received by an instance of the operator must be the same size and type ROI The ROI input must be single rectangle graphic or a vector of rectangle graphics The y coordinates of the rectangular ROI are specified relative the the frame of reference of the stream of images rather than that of the individual input images Outputs Blobs The Blobs output is a vector of Blob objects Each Blob object contains a single run length The run lengths are grouped into contiguous regions Within each group the run lengths are sorted by increasing line number and then by the x location of the end of the run The groups are sorted on the last run length in the group The operator keeps a running line count the run lengths and information derived f
218. d for each vector may be specified using the numToCollect parameter or if the numToCollect parameter is set to zero may be controlled by the Control input The Control input must be supplied each time an input object is received even if the vector size is specified In the case where Control input is used if its value is zero the input object is appended to the vector and collection continues If its value is not zero the collector sends a vector of collected objects and resets itself When the ignoreLast parameter is set to yes the last object fed to the collector is not appended to the outgoing vector The ignoreElement parameter determines whether the input element is added to the current vector or consumed without being added When ignoreElement is set explicitly it is always set to no When used as an input parameter it allows conditional collection of elements based on a boolean value produced in the graph Collector is typically used in conjunction with sequencer where the Control input is provided by the sequencer Parameter Information status Type int Default 0 Values new 0 append 1 ignoreLast Type int Default 0 Values no 0 yes 1 numToCollect Type int Default 0 numCollected Type int Default 0 ignoreElement Type int Default 0 Values no 0 yes 1 elementTypes Type int Default 0 Type choice same 0 different 1 If elementTypes is same only objects of the same type are colle
219. ddition the Intersect output point is colored green if the point is a valid intersection and red if it is not Parameter Information extend Type int Default 0 Values bitwise OR LineO 1 Linel 2 Demo iGraph Operators Geometry convexHull igr C Prototype CorOpRtn cor_lineIntersect CorObj Lined CorObj Linel CorGraphic Intersection int Status int extend Headers include WITHOME h wFit h Libraries WITHOME lib wObj lib WITHOME lib wFit lib mergeLines Merge approximately contiguous line segments 109 Description The mergeLines operator combines line segments from a vector of Line objects into longer line segments The input is a vector of Line data objects typically produced by the geom2lines operator The output is also a vector of Line objects Lines are merged if their angles are within a tolerance specified in degrees by the angle parameter and the endpoints of matching ends are within a distance of each other specified in pixels by the gapSize parameters The operator can also reject line segments smaller than the minLength parameter value Parameter Information gapSize Type int Default 10 minLength Type int Default 20 angle Type float Default 10 Demo iGraph Operators Geometry convexHullL igr C Prototype CorOpRtn cor_mergeLines CorLineVector In CorLineVector Merged int gapSize int minLength float angle Headers include WITHOME h
220. de the range of the output object type the value is clipped to the range If the input is a vector of one of these types the output will be a vector of the same size in which the elements are of the type specified in the type parameter The conversion parameter determines how float or double types are converted to one of the integer output types and how the x and y fields in Real Point objects are converted to the corresponding fields in the Point object Selecting round will set the output value to the nearest integer of the input value floor will set it to the nearest integer less than or equal to the input value and ceiling will set it to the nearest integer greater than or equal to the input value Casting to a String or File type from one of the floating point integer or point including Real Point types results in a text string representation of the value For example an int object with value 10 will be converted to the string JO and a Point object with x 0 y 0 will be converted to the string 0 0 Input String or File type objects can be cast to numeric types When a string is converted to a signed or unsigned integer type it is interpreted using the C programming language convention Strings starting with a 0 are interpreted as octal numbers and strings starting with Ox or OX are interpreted as hexadecimal numbers For example the strings 12 014 and OxC are all cast to the integer 12 If the entire str
221. determined by the range of input type 256 elements for ubyte and byte 65 536 elements for ushort or short For the unsigned types ubyte and ushort the vector element indexes correspond directly to the LUT range values so the first element contains f 0 and the i th element contains f i For the signed types byte and short the vector element indexes correspond to the LUT range values offset so that the minimum pixel value appears in the first vector element For example for a byte image the first element contains f 128 and the i th element contains f i 128 The outputType parameter determines the type of the elements of the output vector which in turn determines the type of the output image of the applyLut operator It may be set to one of the integer or float image types ubyte byte ushort short or float The output values are calculated using floating point arithmetic and if necessary converted to integer values as specified by the conversion parameter Fractional values of 0 5 are always rounded up in the round mode For the integer types values greater than the maximum value in the image type range are set to the 23 maximum value in the range and values less than the minimum value in the range are set to the minimum Many of the common mapping functions can be constructed Refer to the table below for a list of the most useful Function Equation k A digital negative A k A where k maximum pixel
222. dth and height parameters control the size of the output image When a parameter is set to a non zero value the corresponding dimension of the output image is set to the parameter value from a minimum of the corresponding dimension of the smallest scaled output image to a maximum of the corresponding dimension in the largest scaled image When a parameter is set to zero the corresponding output dimension is the size of the scaled image in that dimension When either of these parameters is set to zero the resulting image vector will contain images of different sizes and cannot be used directly as input to the patInit operator The width and height parameters may be used to fix the size of the images in the output vector Pixels in a fixed size output image that are outside the boundaries of the scaled input image are set to the average of the four corner pixels of the input image Parameter Information minScale Type float Default 1 maxScale Type float Default 1 scaleIncr Type float Default 0 1 width Type int Default 0 height Type int Default 0 C Prototype CorOpRtn cor_patScale CorImage In CorImageVector Scaled CorFloatVector Params float minScale float maxScale float scaleIncr int width int height Headers include WITHOME h wSearch h Libraries WITHOME lib wObj lib WITHOME lib wSearch lib patScaleRotate Rotate and scale a pattern image through ranges of values
223. duces an unsigned 8 bit image for gray scale files The operator produces a WiT RGB image for color images The operator can handle JPEG color images in RGB YCbCr CMYK and YCCK formats Other formats will be passed without conversion with the channel pixel values placed in the R G and B channels of the output image In the Raw mode binary files representing color greyscale floating point and complex data images may be read The type and size of the image must be known The type is specified using the rawType parameter In the case of color images the raw data is represented by three files Each file contains the raw data for one of the three color channels red green or blue Each file is named by the filename parameter suffixed with either the letter r g or b For example if the filename is given as xx then the files expected are xx r xx g and xx b Each raw data file is organized in row column order given the width and height parameters Floating point images are single precision and complex images are organized as real complex single precision float pairs The offset parameter is used to skip the given number of bytes into the file prior to extracting the data to accomodate header information Parameter Information filename Type CorFile Default sample wit type Type int Default 0 Values Auto 0 GIF obsolete 1 PCX 2 PBM 3 SUN 4 XBM 5 TIFF 6 BMP 7 JPEG 8 WiT 9 Raw 10 JPEG2000 JP2 11
224. e Headers include WITHOME h wMatrix h Libraries WITHOME lib wObj lib WITHOME lib wMatrix lib SmartOCR Character and text recognition ocrBin Binarization using global thresholding s ABE ye 233 Description The ocrBin operator binarizes a text image based on either a supplied threshold or a threshold calculated from the pixel values in the image Inputs In The input image may be any integer image type or a float image Outputs Out The operator produces a binary unsigned 8 bit image with white text on a black background Threshold The threshold value used to binarize the image is output as a float value This value may be used as an input parameter to subsequent applications of the ocrBin operator to maintain a consistent threshold level Parameters background The background parameter is set to light if the input image has a light background with dark text or to dark if the input image has a dark background with light text Image pixels with a value below the threshold value are set true non zero and those above the threshold as set false zero if the background has been specified as light The opposite is true if the background has been specified as dark calcThresh If the caleThresh parameter is set to No the threshold supplied in the threshold parameter is applied to the image directly If the calcThresh parameter is set to Yes the threshold is calculated from the pixel value
225. e WITHOME h wStats h Libraries WITHOME lib wObj lib WITHOME lib wStats lib getSlopeMaxima Find local slope maxima in a vector Description The getSlopeMaxima operator finds local maxima and or minima in a vector of scalar or point type values The Profile input may be any integer or real scalar type vector or an integer or real point vector The direction parameter determines if the operator looks for rising edges falling edges or both 200 The Slope output is a vector that contains the calculated slope values For scalar inputs this is a float vector For point inputs this is a real point vector that contains the slope values in the y field and the x values from the Profile input the x field The SlopeMax output is a vector of real point values that correspond to local slope extrema in the input that have an absolute value greater than the threshold parameter value The x fields correspond to the position of the slope maximum minimum and the y fields correspond to the maximum minimum slope value The SlopeStart and SlopeEnd outputs are vectors of real points that contain the location in the x field and profile value in the y field of the start and end of a continuous slope region in the profile The start and end of the slope regions are defined as the zero crossing or local exteme value in the slope nearest to the slope maximum minimum location The Slope output is a vector that contains the calculated slo
226. e a LUIN gt Description The getCode128 operator accepts a binary or grayscale image and one or more rectangular regions of interest ROIs then searches the ROIs for Code128 barcodes If a barcode is found in a region it is decoded to a string Inputs Target The input image must be an unsigned 8 bit grayscale image with the barcode represented as dark bars on a light background Roi A single ROI may be specified by a rectangular Graphic or Geometry object Multiple ROIs can be specified by a vector of Graphic or Geometry objects Outputs Decoded If the ROI is a single object then the output is a String object If a valid barcode is found the output string contains the characters represented by the barcode If no valid barcode is found an empty string is output If the ROI is a vector of rectangles then the operator seeks for a barcode in each ROI The output is a vector of String objects If a 223 valid barcode is found in a ROI the corresponding element of the output vector contains the decoded string If no valid barcode is found in a ROI the corresponding output vector element is the empty string Parameters noise The noise parameter specifies the minimum low to high or high to low change in pixel value necessary to constitute the edge of bar Parameter Information noise Type int Default 5 Demo iGraph Operators Matrix Code128 igr C Prototype CorOpRtn cor_getCodel128 CorImage Target
227. e int Default 25 threshFactor Type float Default 0 7 C Prototype CorOpRtn cor_ocrLocalBin CorImage In CorImage Out int background int width int height float threshFactor Headers include WITHOME h wOcr h Libraries WITHOME lib wObj lib WITHOME lib wOcr lib ocrMakeFont Extract font characters from an image Description The ocrMakeF ont operator adds characters from an example image to an OcrFont object The regions corresponding to new characters in the example image are specified by graphic ROIs and the operator s parameters The input image must an integer or float gray scale image type The ROI input must be either a Graphic or Geom object or a vector of Graphic or Geom objects The graphics may be rectangle or point type The ROI is a rectangle whose size may be specified by the width and height parameters The ROI rectangle is centered at the point location for point graphics or at the center of the rectangle for rectangle graphics Note that in this case the rectangle graphic is not used directly as the ROI If the width height parameter is set to 0 and the input graphic is a rectangle the width height of the ROI will be set to the rectangle s width height The ROI will be the input graphic rectangle if both width and height are set to 0 The region of the image specied by the ROI is extracted and used as a model image for the font If the ROI extends outside the input
228. e ocrGrayInit operator to produce font models for the ocrGrayRun operator Parameter Information labels Type CorVector C Prototype CorOpRtn cor_ocrSelectFont CorOcrFont InFont CorOcrFont Selected CorOcrFont Remaining CorStringVector labels Headers include WITHOME h wOcr h Libraries WITHOME lib wObj lib WITHOME lib wOcr lib SmartSearch Fast and accurate pattern search 248 patFind1 Find a pattern or patterns Mt Description The patFind1 operator locates the best match or matches to one or more patterns in a rectangular region of interest ROI of the target image The inputs to the operator are a gray scale target image a rectangle graphic specifying the ROI and a pattern model which may use tiled pattern images and or binary masks Matches between the patterns and the image are evaluated by a statistical technique using pixel grayscale values A two level strategy is used to speed up the search for matches The operator produces the location match score and pattern index number of the match or matches meeting criteria specified by the operator s parameters Inputs Target The target image can be any gray scale WiT image an unsigned or signed eight or sixteen bit integer image or a float image The operator does not accept color or complex images ROI The ROI must be a rectangle type Graphic object The operator searches locations in the target image for which the pa
229. e after it has been reached Otherwise the counter is incremented indefinitely Counters are referenced by the name parameter which refers to a global name See chapter Operators Special in User s Manual for details See also counter Reset counterStep Parameter Information name Type CorString Default counter Demo iGraph Operators clearObjs data2sync Convert a data token to a sync token aty Description The data2sync operator takes an input object of any type and produces a sync token Demo iGraph Operators clearObjs delay Delay for a specified time Description The delay operator delays igraph execution for a number of milliseconds specified by its msec parameter Parameter Information msec Type int Default 1000 Demo iGraph Operators clearObjs for Repeat a set of operations a specified number of times Description The for operator is useful for executing a set of operators a fixed number of iterations It has one standard integer parameter count and one special parameter currentCount The currentCount parameter is initially set to zero and cannot be set by the user Every time a token arrives at the input port currentCount is incremented by one If the resulting value of currentCount is less than the value of count the token is passed to the top branch labelled Loop which most often would have a sequence of operators eventually sending a token back to the inp
230. e depending on view point Cutplane Bring up the image properties panel See chapter Operators Libraries in User s Manual for details Image properties The operation of the Surface ramp control is similar to the colormap ramp for images except that the color changes affect surface color before shading is applied rather than affecting the rendered image colors directly Opacity Ramp The Opacity ramp control Surface Ramp arrows can be used to compress and shift the selected opacity map Transparent values are represented by black opaque by white and semi transparent by gray Parameter Information name Type CorString Default volume Placement Type int Default 0 Values bitwise OR Fix position 1 Max size 2 X Type int Default 0 Y Type int Default 0 MaxW Type int Default 0 MaxH Type int Default 0 Edges The Edges library supports the extraction of edges and contours from images The library includes operators that find point line or circular edges based on image profiles and operators that extract contours from grayscale or binary images getCircularXEdges Fits a circle to the edge points Description The getCircularXEdges operator finds edge points near a nominal circle or circular arc and fits a circle to the points The operator s inputs are In a grayscale image and NomCircle a circle or arc type Graphic or Geometry object The opera
231. e getField operator may be used to extract any field given its name for use with operators like conditional to filter certain object attributes Care must be taken to insure that the image used by getBlobs to generate the blob vector is also used by getPerimeter to generate the perimeter data The connected parameter can be used to specify whether the perimeter pixels should be eight connected where diagonal neighbours are considered to be touching or four connected where they are not The type parameter is used to to specify whether the input image is considered a binary image in which case non zero regions will be traced or a labeled image in which case equal valued non zero regions will be traced In general the connected and type parameters should be set to the same values as the connected and type parameters on the getBlobs or getBlobFeatures operator that was used to generate the input Blob feature vector If they are not the operator may have trouble correctly tracing the perimeter Parameter Information connected Type int Default 1 Values 4 0 8 1 type Type int Default 0 Values binary 0 labeled 1 Demo iGraph Operators Blobs Holes C Prototype CorOpRtn cor_getPerimeter CorFeatureVector In CorImage BinaryImage CorFeatureVector Out int connected int type Headers include WITHOME h wBlob h Libraries WITHOME lib wObj lib WITHOME lib wBlob lib getWarpedFeatures Gener
232. e loOffset and hiOffset parameter values are simply added to the mean value to determine the low and high threshold values In the other modes the parameters are used as multiplying factors to determine the threshold offset values In the mean min mode the difference between the image mean pixel value and the image mimimum pixel value is multiplied by the parameter values and the resulting offset values are added to the image mean In the mean min mode the difference between the image maximum and the image mean is multiplied by the parameter values to determine the offsets In the standard deviation mode the offsets are determined by multiplying the standard deviation of the image pixel values by the parameter values The low and high threshold values calculated for the image are output at the LoThresh and HiThresh outputs See the thresh operator for a description of the invert levels and result parameters and the output images that result from various combinations of these parameters Parameter Information offsetMethod Type int Default 0 Values absolute 0 mean min 1 max mean 2 standard deviation 3 loOffset Type float Default 0 hiOffset Type float Default 0 invert Type int Default 0 Values no 0 yes 1 levels Type int Default 1 Values double 0 single 1 result Type int Default 1 Values semi threshold 0 binary 1 Demo iGraph Operators Segmentation AdaptThresh 190 C Prototype
233. e local pixel value entropy Description The entropy operator produces an image made up of the entropy of the input image pixel values in a specified neighbourhood around each image pixel The entropy is calculated by summing the product pi In p over all pixel values i where p i is the fraction of pixels in the neighbourhood with value i The operator accepts a gray scale integer or float image input and produces a float image as output The size of the neighbourhood over which the entropy is calculated is specified by the width and height parameters Parameter Information width Type int Default 5 height Type int Default 5 Demo iGraph Operators Filter Entropy Operators Filter V arianceTexture C Prototype CorOpRtn cor_entropy CorImage In CorImage Out int width int height Headers include WITHOME h wFilter h Libraries WITHOME lib wObj lib WITHOME lib wFilter lib gauss Two dimensional gaussian smoothing Description The gauss operator performs a two dimensional smoothing of the input image using a discrete approximation to a gaussian kernel The effect of the smoothing operation is to blur the image reducing the sharpness of edges and fine detail in the image and reducing the effects of high frequency noise An analytic gaussian filter more closely approximates a perfect low pass filter than the moving average filter used in the lopass2d operator while avoiding the ringi
234. e operations The output parameter can be used to select either a grayscale output image in which the output image represents an approximation to the gradient magnitude or a binary output image in which case the threshold is applied to the gradient magnitude image to extract edges The edgeDetect operator uses the maximum of eight directional derivatives calculated using either 81 Prewitt or Sobel kernels as specified by the type parameter to approximate the gradient magnitude This approach is somewhat faster than the method used to produce the magnitude output of the grad operator This operator could be executed in two modes The user may choose the mode by using the optimization parameter If the parameter value is speed then execution goes way faster otherwise operator delivers more precise results Parameter Information type Type int Default 0 Values Prewitt 0 Sobel 1 output Type int Default 0 Values Binary 0 Gray Level 1 threshold Type float Default 0 optimization Type int Default 1 Values preciseness 0 speed 1 Demo iGraph Operators Filters conv2d igr C Prototype CorOpRtn cor_edgeDetect CorImage In CorImage Out int type int output float threshold int optimization Headers include WITHOME h wFilter h Libraries WITHOME lib wObj lib WITHOME lib wFilter lib 82 edgeDirection Gradient magnitude and direction filter Eo o Description
235. e parameter The angle is given in degrees where positive angles shear the image clockwise Shearing constrains an image rotation to a single degree of freedom resulting in a stretching of the image along the y axis about the given angle The range of angle is 90 90 Parameter Information angle Type float Default 10 Demo iGraph Operators Fit polar igr C Prototype CorOpRtn cor_shearY CorImage In CorImage Out float angle Headers include WITHOME h wGeom h Libraries WITHOME lib wObj lib WITHOME lib wGeom lib warp Apply perspective transform matrix to object imf Tip Description The warp operator applies a 3x3 spatial transformation matrix to an image or graphical object 132 Images are mapped to the transformed space using bilinear interpolation and are clipped to the size of the original input image The operator will also accept or single objects of the types Point Real point Graphic Geometry Line or Edge The length and angle fields of Line objects are adjusted to match the transformed endpoints The direction slope and level fields of Edge objects remain unchanged Circles and rectangles are converted to polygons before transformation and the output objects are polygons The starting position of text objects is transformed but the actual text is not distorted The warp operator is usually used with a matrix generated by the getWarpTransform operator Demo iGraph Operators
236. e the robustness of matching of partially occluded or damaged target images Tiled model images can also be used to maintain precision in large greater than 256x256 model images by suppressing the automatic scaling that is applied to large model images The operator can also incorporate binary mask images into the pattern model Only pixels in the model and target images that correspond to non zero pixels in the mask image will be used in the calculation of matching scores Inputs ModelImages The input must be a vector of image vectors All images within an image vector must be the same size but the image size may vary among the image vectors The images must all be gray scale image of the same image type The input images and the resulting pattern models in either dimension may represent variations on a pattern or different basic patterns A typical example is an input vector in which each image vector represents a different scaling of a pattern image and each image within a vector represents a different rotation of the scaled image The patRotateScale operator produces a vector of image vectors in this format from a single image When applicable the images should be more closely related within each image vector than they are across the vector of vectors For example each image vector in the top level vector might represent a different character while the images within a vector represent various 257 Outputs forms
237. e third is the minimum rotation angle and the fourth is the angle increment Parameters minScale maxScale and scaleIncr The input image is scaled by factors from minScale up to the largest angle less than or equal to maxScale in increments of scaleIncr The value of minScale must be less than or equal to the value of maxScale and scaleIncr must be greater than zero startAngle endAngle and angleIncr Each scaled image is rotated by angles from startAngle up to the largest angle less than or equal to endAngle in increments of angleIncr The value of 268 startAngle must be less than or equal to the value of endAngle and angleIncr must greater than zero width and height The width and height parameters control the size of the images in the output vectors If the width or height parameter is set to a non zero value the corresponding dimension of the output image is set to the parameter value scaled by the scaling factor for that image up to a maximum of the corresponding dimension in the scaled input image When a parameter is set to zero the corresponding output dimension is the same as the dimension in the scaled input image The image size is the same within each image vector in the output since all rotations of a scaled image are the same size The image size will vary among the image vectors in the output with the relative size of the images in a vector proportional to the scaling factor associated with that vector
238. e value specified by the UnitsPerPixel parameter to convert them into an appropriate coordinate system and may have an associated unit specified by the UnitName parameter For example you can specify that each pixel represents 2 5 microns by setting the UnitsPerPixel parameter to 2 5 and the UnitName to micron Then all subsequent measurements will be scaled accordingly For example a line 10 pixels long will be reported to be 25 units long Measured values related to the pixel values include the mean standard deviation minimum and maximum of the pixel values Pixel values on the perimeter can also be plotted on a popup graph Properties Main panel name Name of the window used to display the data type The initial graphic type to be entered on the image UnitName The unit name used to report measurements on the pop up panel Can be any valid string UnitsPerPixel A floating point scale factor that specifies the number of linear units each pixel represents This scale factor is used to scale all reported measurements Placement sub panel Placement If Fix position is checked values of the X and Y properties are used to position the data window Otherwise the current automatic placement rules are used X The X position of the data window if Placement is set to Fix position Y The Y position of the data window if Placement is set to Fix position Property Panel x Type Lina Linde
239. ected Type int Default 1 Values 4 0 8 1 type Type int Default 0 Values binary 0 labeled 1 maxBlobHeight Type int Default 0 reset Type int Default 0 Values No 0 Yes 1 resetBlobld Type int Default 0 Values No 0 Yes 1 C Prototype CorOpRtn cor_webGetBlobs CorImage BinaryImage CorBlobVector Blobs int connected int type int maxBlobHeight int reset int resetBlobId Headers include WITHOME h wWeb h Libraries WITHOME lib wObj lib WITHOME lib wWeb lib webGetBlobsRoi Run length encode blobs in ROIs in a stream of line scan images Bag z oe f Description The webGetBlobsRoi operator performs pixel connectivity analysis within rectangular regions of 278 interest on a stream of binary or labeled images The operator extracts run lengths from the image and outputs them in a vector in which run lengths from each contiguous set of pixels in the image are grouped together and given a distinct id number Images in the sequence are treated as if the last line of the previous image is adjacent to the first line of the current image and the first line of the next image is adjacent to the last line of the current image allowing blobs to extend across more than one image Blobs are output when they do not reach the bottom of the current image and so cannot extend into the next image in the sequence Blobs that reach the bottom row of the current image are usually save
240. ected in igraph of course e Out must be connected to calculate Count e The n parameter is not supported Its use will cause unpredictable results e Ifthe input object type is not a string sscan will not work Strings in WiT must start with a character ie the input string 44 hello world will be considered to be an integer type and cause sscan to fail The case where the first is included in the set is not supported The case 4 where the first is included in the set is not supported e Width specifiers single byte wide are not supported for c C s and S e Format specifiers xi and oi are unsupported use x and o to get hex and octal e The 164 prefix for integers is not supported by WiT Parameter Information format Type CorString Demo iGraph Operators Blobs GetBlobFeatures igr Operators Blobs GetBlobFeaturesMasked igr Operators Blobs SelectFeatures igr C Prototype CorOpRtn cor_sscan CorObj In int Count CorVector Out CorString format Headers include WITHOME h wConvert h Libraries WITHOME lib wObj lib WITHOME lib wConvert lib strLength Get string length Tp Description The strLength operator outputs the length of the input string Demo iGraph Operators Blobs GetBlobFeatures igr Operators Blobs GetBlobFeaturesMasked igr Operators Blobs SelectFeatures igr C Prototype CorOpRtn cor_strLength CorString In int Lengt
241. ecting them and the distance between points in pixels as text These may be displayed on an image using the overlayData operator 140 The color of the graphical objects can be selected from a number of color combinations with the color parameter The points and text are printed in the main color the connecting lines are printed in the secondary color The font fontSize and fontStyle parameters control the output text If the font parameter is set to no text no text is produced Parameter Information color Type int Default 0 Values red yellow 0 yellow green 1 aqua blue 2 white yellow 3 black blue 4 font Type int Default 0 Values helvetica 0 times 1 courier 2 no text 3 fontSize Type int Default 12 fontStyle Type int Default 0 Values normal 0 bold 1 italic 2 bold italic 3 Demo iGraph Operators Measurement CircleCenters Operators Segmentation Watershed C Prototype CorOpRtn cor_plotDist CorVector Points CorGraphicVector Out int color int font int fontSize int fontStyle Headers include WITHOME h wMeas h Libraries WITHOME lib wObj lib WITHOME lib wMeas lib Morphology The Morphology library supports morphological and other non linear neighbourhood operations on images The library includes binary and grayscale erosion dilation and related morphological operators and median and rank operators bmedian Binary median filter m Des
242. ector Default 1 1 1 scale Type int Default 3 Demo iGraph Operators Filters conv1d igr C Prototype CorOpRtn cor_conv1d CorImage In CorImage Out CorIntVector kernel int scale CorOpRtn cor_convld_consume CorImage In CorIntVector kernel int scale Headers include WITHOME h wFilter h Libraries WITHOME lib wObj lib WITHOME lib wFilter lib conv2d General two dimensional convolution D Description The conv2d operator applies two dimensional convolution by a user specified kernel to the input image With an appropriate choice of kernel values the conv2d operator can perform such operations as smoothing edge or peak enhancement position shifting on the image The two dimensional convolution kernel is an integer array specified by the kernel parameter The kernel size and weight values can be changed with the array editor by clicking on the modify button for the kernel parameter The convolution operator functions with odd or even kernel sizes The user may specify a factor using the scale parameter by which all values in the convolved image will be divided This allows the specification of normalized filters of the type used in the gauss operator This operator could be executed in two modes The user may choose the mode by using the optimization parameter If the parameter value is speed then execution goes way faster otherwise operator delivers more precise results
243. ector can then be used to search for scaled patterns using the patFind operator Inputs In The input image can be a gray scale or color image complex images are not currently supported by this operator Outputs Scaled The output is a vector of scaled versions C Prototype CorOpRtn cor_patRotate CorImage In 266 of the input image The number of scaled images the scale factors and the image sizes are controlled by the parameters as described below The images in the output vector are of the same type as the input image Note that vectors color images produced by this operator cannot be used as input to the patInit operator The scaled images are produced by pixel sampling relative to the center pixel in the image with no interpolation or smoothing Params The Params output is a four element float vector that contains the transformation parameters applied to the images in output vector The first element is the minimum scale factor the second is the scale increment the third is the minimum rotation angle and the fourth is the angle increment For the patScale operator the last two values are always 0 Parameters minScale maxScale and scaleIncr The input image is scaled by factors from minScale up to the largest value less than or equal to maxScale in increments of scaleIncr The value of minScale must be less than or equal to the value of maxScale and scaleIncr must be greater than zero width and height The wi
244. ector of templates is used if a single template image is used all elements in this vector are zero The operator requires three inputs the Target input is an integer or float image the Template input is an integer or float image or vector of images and the ROI input is a graphic object or vector of graphic objects that controls the search strategy A template is formed from the ROI of the Template input if the input is a single image or from the ROI of each image if the input is a vector of images The template is compared to regions in the target corresponding to each pixel in the input graphic object or objects The template image is aligned so that the specified target pixels correspond to the center of the template if the result in the Result sub panel parameter is set to center or bounding box or to upper left corner of the template if result is set to upper left For point graphic objects the operator will also search concentric squares of pixels around the point out to the distance specified by the searchRange parameter A comparison will 96 be made for a pixel only if the entire template falls within the target image To search the entire image the bounding rectangle may be obtained from the target image using the getRect operator and apply it to the lower input port If a vector of template images is specified the entire search region in the target image will be scanned for a template image before the next template im
245. ed Operations may result in values which fall outside the range of the output type out of range values are clipped to the maximum or minimum range value To avoid clipping the input images can be cast a higher precision type using the castImage or matchSize operators Color input images may be combined with any type of grayscale image or with a color image of the same type RGB with RGB HSV with HSV and Yuv with Yuv If both input images are color images the operations are applied channel by channel If one of the images is a grayscale image its pixel values are applied to each channel of the color image The output image will be the same type as the color input image The output image will be the same size as the A input image The output image ROI will be the 20 intersection of the ROIs of the A and B images The specified operation will only be applied to pixels in the output image ROI Supported operations on input images A and B are A B A B A B A B OR bitwise OR of A and B AND bitwise AND of A and B XOR bitwise XOR of A and B src use B where B 0 else use A overlay use B if A amp overlayMask and B 0 else use A expr utilize expression to determine result abs A B absolute value of difference between A and B min minimum of A and B max maximum of A and B Each pixel of the output is determined by performing the indicated operation on the corresponding pixels in image A and image
246. edge sumx Sum of all x coordinates sumy Sum of all y coordinates sumxsq Sum of all x coordinates squared sumysq Sum of all y coordinates squared sumxy Sum of all x coordinates y coordinates perimx X coordinate of a pixel on the blob perimeter lower right corner perimy _ Y coordinate of a pixel on the blob perimeter lower right corner xc X coordinate of blob centroid sumx n ye Y coordinate of blob centroid sumy n angle Angle of best fit ellipse major axis clockwise from x axis major Length of best fit ellipse major axis minor Length of best fit ellipse minor axis axratio minor major xdiff xmax xmin 1 ydiff ymax ymin 1 boxarea__ xdiff ydiff bxaratio npixels boxarea perim Perimeter data for blob must run the getPerimeter operator to fill in this structure The getField operator may be used to extract any field given its name for use with operators like conditional to filter certain object attributes Demo iGraph Operators Blobs Holes C Prototype CorOpRtn cor_getFeatures CorBlobVector Blobs CorFeatureVector Features Headers include WITHOME h wBlob h Libraries WITHOME lib wObj lib WITHOME lib wBlob lib getPerimeter Find perimeter given a Blob feature ETTI K Sea Description The getPerimeter operator determines the perimeter related data corresponding to the input Blob feature vector and its associated input image The resulting Blob feature vector is a du
247. educes the width of the thick edges in images processed by grad edgeDirection or edgeDetect operators Parameter threshold is applied to the image to eliminate weak edges Parameter Information threshold Type float 91 Demo iGraph Operators Filter Entropy Operators Filter V ariance C Prototype CorOpRtn cor_thinkEdges CorImage Edges CorImage ThinEdges float threshold CorOpRtn cor_thinEdges_consume CorImage Edges float threshold Headers include WITHOME h wFilter h Libraries WITHOME lib wObj lib WITHOME lib wFilter lib variance Compute local image variance g i Description The variance operator calculates the mean and the variance or standard deviation of the image pixel values in a specified neighbourhood around each image pixel The operator accepts a gray scale integer or float image input and produces float images as output The output parameter is used to specify whether the operator produces the variance or standard deviation image as output The size of the neighbourhood over which the variance is calculated is specified using the width and height parameters 92 Parameter Information output Type int Default 0 Values variance 0 std dev 1 width Type int Default 5 height Type int Default 5 Demo iGraph Operators Filter Variance Operators Filter VarianceTexture C Prototype CorOpRtn cor_variance CorImage In CorImage Mean Cor
248. eeccscesseseseeseesesseeseeseeeeneeeeaes 302 acqWriteLut 303 acqZOom 303 acquire oossoo 303 acquireStamped vi Standard Blobs The Blobs library provides blob analysis capabilities The library includes operators that extract and process blobs which are connected components in binary or labeled images and a set of features of the blobs such as area and centroid andBlobs Feature AND of blobs Description The AndBlobs operator performs an asymmetric AND of two binary images represented as run length encoded blobs The result of the asymmetric AND of image A with image B is an image containing those blobs from A which have at least one pixel in common with one or more blobs from B This AND operation is not symmetric so A AndBlobs B is not equal to B AndBlobs A The AndBlobs operator is applied to vectors of blobs run length encoded regions of contiguous non zero pixels extracted from images using the getBlobFeatures operator The output at the And port is a vector of blobs containing those blobs from the vector on the A input that have at least one pixel in common with one or more blobs in the vector on the B input The output at the NAnd port is a vector of those blobs from A that do not have any pixels in the blobs from B The output blob vector can be converted to a feature vector with the getFeatures operator or into an image with the labelBlobs operator Demo iGraph Operat
249. egion it is decoded to a string Inputs Target The input image must be an unsigned 8 bit grayscale image with the barcode represented as dark bars on a light background Roi A single ROI may be specified by a rectangular Graphic or Geometry object Multiple ROIs can be specified by a vector of Graphic or Geometry objects Outputs Decoded If the ROI is a single object then the output is a String object If a valid barcode is found the output string contains the characters represented by the barcode If no valid barcode is found an empty string is output If the ROI is a vector of rectangles then the operator seeks for a barcode in each ROI The output is a vector of String objects If a valid barcode is found in a ROI the corresponding element of the output vector contains the decoded string If no valid barcode is found in a ROI the corresponding output vector element is the empty string Parameters noise The noise parameter specifies the minimum low to high or high to low change in pixel value necessary to constitute the edge of bar Parameter Information noise Type int Default 5 Demo iGraph Operators Matrix UPSa igr C Prototype CorOpRtn cor_getUPCa CorImage Target CorObj ROI CorObj Decoded int noise Headers include WITHOME h wMatrix h 231 Libraries WITHOME lib wObj lib WITHOME lib wMatrix lib getUPCe Finds and decodes UPCe bar code S Description The get
250. eground pixels are set to zero and the background pixels contain the distance to the nearest foreground pixel Distance maps can also be produced directly from Graphic objects using the grDistMap operator Parameter Information kernelType Type int Default 0 Values Euclidean 3 x 3 0 Euclidean 5 x 5 1 City Block 2 Chess Board 3 distFrom Type int Default 0 Values background 0 foreground 1 Demo iGraph Operators Measurement CircleCenters Operators Segmentation Watershed C Prototype CorOpRtn cor_distMap CorImage In CorImage Out int kernelType int distFrom Headers include WITHOME h wMeas h Libraries WITHOME lib wObj lib WITHOME lib wMeas lib distPtToLine Measure the distance from point s to a line oe an Description The distPtToLine operator measures the distance from a point or points to a line The Pt input may be a single point represented by a point type geometric object Geometry Graphic or Edge object type or a Point or Real Point object or it may be a number of points represented by a vector of such objects The Lineinput must a line type geometric object Geometry Graphic or Line object type For a single input point the Dist output is a float value that represents the shortest distance from the input point to the line and the NearestPt output is a Graphic object representing the point on the line nearest to the input point For a vector of input points
251. ell PatModel CorFpointVector Locations CorFloatVector Scores CorIntVector Indexes int scale int range float rejectThresh float acceptThresh float minContrast int maxTilesOmitted int localSearch int maxMatches float relThresh int selectMode int xVicinity int yVicinity int applyPatterns int patSampling CorIntVector patCenter int patRange Headers include WITHOME h wSearch h Libraries WITHOME lib wObj lib WITHOME lib wSearch lib patRoi1 Find a pattern in a rectangular region of interest E a it Description The patRoi operator locates the best match or matches to one or more patterns in a rectangular region of interest ROI of the target image Matches between the patterns and the image are evaluated by a statistical technique using pixel grayscale values The inputs to the operator are a gray scale target image a rectangle graphic specifying the ROI and a vector of pattern models which may use tiled pattern images and or binary masks The operator outputs the location match value and pattern element number of the match or matches meeting criteria specified by the operator s parameters The operator can search at reduced resolution to increase speed Inputs Target The target image can be any gray scale WiT image an unsigned or signed eight or sixteen bit integer image or a float image The operator does not accept color or complex images ROI The ROI must be a rec
252. els apart in x and yVicinity pixels apart in y when the selectMode parameter is set to Location or Both The MultiplePatterns Sub Panel The MultiplePatterns sub panel includes parameters that control the output when multiple pattern models are used applyPatterns The applyPatterns parameter is only meaningful when more than one pattern model is applied When the parameter is set to All all patterns in the input pattern vector are compared to the target image in both phases of the search When the parameter is set to Sample the pattern vector is sampled as specified by the patSampling parameter for the candidate search A range of adjacent pattern models can be selected by setting the parameter to Range in which case the elements used are specified by the patCenter and patRange parameters Finally the user can specify arbitrary pattern models by setting this parameter to Select and specifying the patterns using the patIndexes parameter patSampling The patSampling parameter specifies the sample spacing in the pattern vector for the search when applyPatterns is set to Sample patCenter and patRange The patCenter and patRange parameters specify elements selected from the pattern vector for the search when applyPatterns is set to Range A contiguous group of 2 patRange 1 elements centered on element patCenter are used in the search Parameter Information scale rejectThresh minContrast Type int Default 2 Values
253. ements in the kernel array Erosion by a rectangular structuring element is usually faster than erosion by an arbitrary element particularly for larger neighbourhoods One common exception is erosion by 3x3 structuring elements where the arbitrary kernel usually is faster Binary erosion tends to make objects in the binary image smaller separating objects that are touching and removing isolated pixels The kernel either rectangular or arbitrary specifies the structuring element which will control the erosion The center of the kernel is always in row height 2 and column width 2 For each pixel p in the input image if every pixel in the neighbourhood of p as defined by the kernel dimensions is on and the corresponding entry in the kernel is also on then the output pixel corresponding to p will be set Otherwise it will be cleared In other words the kernel defines a mask which is layed over the input image with its center aligned on a pixel at row r and column c The output pixel at r and c will be set if every 143 combination input pixel is set and its corresponding bit in the kernel mask is also set For example oorr row OrRrRrF Oc ornoo Siere G 1 Kernel 011 Output 00000 01000 Oe 8 0 070 2 0 0 0 0 00 0 Output 00000 lo ER o 80 O oE 1 0 0 0 0 0 8 0 lt 00 0 2 Kernel 1 1 1 0 Output 00000 01000 00100 00000 0 0 0 00 In the first example the output is only reduced on the right side of the in
254. en the image does not divide evenly into the specified size or number of subimages In the truncate mode the image is divided into equal size sub images starting in the top left corner Pixels on the right or bottom of the image that do not fit are dropped Behaviour in the fit mode depends on the value of the set parameter When it is set to size any remainder pixels will be formed into a sub image smaller than the specified size In the number mode the size of the last sub image in the row or column will be adjusted to account for the remainder In this case the last subimage may by larger or smaller than the other sub images by up to half the specified number of columns or rows Images decomposed using the getTiles operator can be reassembled using the tileImages operator Parameter Information set Type int Default 0 Values size 0 number 1 width Type int Default 32 height Type int Default 32 nCols Type int Default 8 nRows Type int Default 8 remainder Type int Default 0 Values truncate 0 fit 1 C Prototype CorOpRtn cor_getTiles CorImage In CorVector Out int set int width int height int nCols int nRows int remainder Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib gf Extract a field from input object Description gf performs the same function as getField gf has a small icon which makes it more convenient
255. en the line edges is measured using the intersection of the reference line with one or both line edges as measurement points The operator s inputs are a gray scale image and the reference line The reference line must be a line geometric object Geometry Graphic Line or Edge type object The operator s outputs are an Edge vector containing the two edges found a float value giving the distance between the measurement points on each edge and a Graphic vector containing the two measurement points used to determine the distance The measure parameter is used to specify how the measurement points on the edge lines are determined Edge line EO is the edge line near the first end point of the reference line edge line E is the edge line near the second end point of the reference line point PO is the intersection of the reference line and EO and point PJ is the intersection of the reference line and EZ Setting measure to PO P sets the measurement points to the PO and P intersections Setting the parameter to PO E perp to EO sets the measurement points to PO and the intersection of E and a line through PO perpendicular to EO Setting the parameter to PO E1 shortest sets the measurement points to PO and the point on EO nearest to PO Setting the 134 parameter to MidPoint sets the measurement points to the intersection of the edge lines and a line through the mid point of the reference line at an angle that is midway between perpe
256. er and float type input images and color images in the case optimisation is speed The output image is the same type as the input image Parameter Information width Type int Default 3 height Type int Default 3 optimization Type int Default 1 Values preciseness 0 speed 1 Demo iGraph Operators Filter Entropy Operators Filter Variance C Prototype CorOpRtn cor_lopass2d CorImage In CorImage Out int width int height int optimization CorOpRtn cor_lopass2d_consume CorImage In int width int height int optimization Headers include WITHOME h wEFilter h Libraries WITHOME lib wObj lib WITHOME lib wFilter lib prewitt Prewitt edge enhancement na F Description The prewitt operator performs a convolution of the input image with Prewitt horizontal or vertical 89 kernels The effect of the Prewitt operation is to enhance the horizontal or vertical edges of the image Two different filters are provided A horizontal or vertical edges enhancement filter may be selected using the direction parameter Use of the prewitt operator is equivalent to using conv2d with the following filters and scale parameters Type horizontal Scale 6 1 1 1 0 0 0 ef g Type vertical Scale 6 4 Or al 1 Oo 1 1 0 1 optimisation paramiter defines wether fast or precise version will run The operator accepts integer and float type input images for precise and integer type for
257. er or getPerimeter operators Parameters yOffset 270 The yOffset value is used with the blockHeight parameter value to determine if a cluster should be released when the reset parameter is set to No or is a zero if it is an input parameter Any cluster whose minimum y value would be below the yOffset value if it were extended into the next block will be released If the clustered features are going to be used with images produced by the webWaterfall operator then this parameter should be an input parameter connected to the Offset output of the webWaterfall operator blockHeight The blockHeight parameter is used with the yOffset parameter to control the release of clusters of Features This parameter should be set to the height of the images from which the Features were extracted The operator expects these images to be the same height reset When the reset parameter is set to Yes the operator forces all clusters accumulated to be output If the reset parameter is set to No the operator will output clusters that could extend beyond the yOffset value if grouped with a Feature from the next input vector All other clusters will be held to determine if they will be extended into subsequent images The reset parameter should always be set to the same value as the reset parameter of the webGetBlobs operator that generated the blobs that produced the Features resetld When the resetId parameter is set to No the the Feat
258. erator calculates the distances between all or a subset of the pairs of points in the input point set The input point set may be a vector of Point Real Point or Point Value objects or a vector of point type Graphic or Geometry objects The operator outputs two vectors of vectors the Distances vectors and the Indexes vectors The output vectors contain one vector element for each point in the input vector Each element of the Distances vector if a vector of Real numbers that contains the distances from the point in the corresponding element of the input vector to each other point in the input vector sorted from the smallest to largest The Indexes vector is made up of Integer vectors in which the elements are the indexes of the points in the input vector that produced the distance in the corresponding element of the Distances vector The maximum number of distances calculated for each point can be specified using the maxDistances parameter When this parameter is set to a value greater than zero at most that number of distances and indexes will be included in each element of the output vector When this parameter is set to zero all distances will be produced 135 Parameter Information maxDistances Type int Default 0 C Prototype CorOpRtn cor_distA11PtToPt CorObj Points CorVectorVector Distances CorVectorVector Indexes int maxDistances Headers include WITHOME h wMeas h Libraries WITHOME lib w
259. erator performs an iterative least squares fit to fit a Circle to the input points The operator s input may be a vector of Point Fpoint PtValue Graphic Geometry or Edge objects or a single Graphic or Geometry object If the input is a vector of Graphic Geom or Edge object each object in the vector must of point type There must be at least three points in the input If 101 the input is a Graphic or Geometry object it may be any type except circle and text In this case the circle will be fit to the vertices treated as independent points If the input is a vector of PtValue objects the value field of the elements of the vector will be used as a weight value for a weighted least squares fit Negative weight values are ignored The operator s main output is a single circle Graphic object The number of iterations required is to fit the points is also produced The if autoInit parameter is set to Yes the operator will use the centroid of the point positions as an initial guess at the circle s center position If it is set to No the user must supply an initial center position using the initCenter parameter The maxIterations parameter can be used to limit the number of iterations the operator will perform Parameter Information autolnit Type int Default 1 Values No 0 Yes 1 initCenter Type CorFpoint maxlterations Type int Default 256 Demo iGraph Operators Geometry convexHullL igr C Prototype CorOpRtn cor_
260. erence manual for details zoom The density of the grid squares in pixels Pixels are grouped into square tiles and pixels within the same tile are drawn with the average color of all the pixels in that tile If this value is 0 the tile sizes will be calculated automatically If this value is 1 the surface is not shown only the wireframe is shown grid wireframe The density of the wireframe in pixels A value of 1 means that the wireframe will have the same density of the surface tiles grid Otherwise the grid can be denser than the wireframe but not the other way If the surface is not shown grid is 1 then the wireframe edges will follow each pixel value The color to be used for the wireframe The value should be an RGB triplet E g 255 0 0 means red wireColor Le PEE Placement sub panel Placement If Fix position is checked the parameters X and Y are used to position the window Otherwise the current automatic window placement strategy is used to position the window lt 67 X The X horizontal position of the window on the screen if Placement has Fix position checked Y The Y vertical position of the window on the screen if Placement has Fix position checked width The width of the window used to display the surface plot height The height of the window used to display the surface plot Image sub panel Same as Image sub panel for the display operat
261. ernel int optimization Headers include WITHOME h wMorpho h Libraries WITHOME lib wObj lib WITHOME lib wMorpho lib morphClose Morphological closing a ee E Description The morphClose operator performs binary or gray scale morphological closing on images Binary opening tends to fill small holes or gaps in larger objects without making the larger objects larger Binary closing is implemented by applying binary dilation followed by binary erosion to the image Gray scale closing tends to increase reduce pixel values in small regions with low values relative to the surrounding area without increasing the value at the edges of large low valued regions Gray scale closing is implemented by applying gray scale dilation followed by gray scale erosion to the image This operator could be executed in two modes The user may choose the mode by using the optimization parameter If the parameter value is speed then execution goes way faster otherwise operator delivers more precise results Binary or gray scale closing is selected using the type parameter The input for binary closing must be an 8 bit unsigned image The result is a binary 8 bit unsigned image For gray scale closing the input may be any integer or float gray scale image or color in the case optimisation is speed The output image is a gray scale image of the same type as the input image The kernelType parameter is used to select the type of s
262. es WITHOME lib wObj lib WITHOME lib wVideo lib acquire Acquire an image from a frame grabber B Description The acquire operator grabs an image from the currently installed frame grabber Typically the Frame Grabber tool in WiT is used to configure the frame grabber for acquisition then the acquire operator is used to grab images in an igraph The frames parameter determines the number of images to grab If frames is set to one then a single WiT image is output If frames is greater than one then a sequence of images is grabbed and output as a vector of WiT images In this case the delay parameter sets the number of milliseconds between image grabs in the sequence The bottom output of the operator reports the number of missed frames since the last acquire This output is meaningful only when acqStart has been used to start continuous acquisition it is 0 otherwise This 303 allows tracking the number of missed frames in the event the processing load on the CPU s prevents the acquire operator from being executed often enough to keep up with the frame rate Parameter Information frames Type int Default 1 delay Type int Default 0 C Prototype CorOpRtn acquire CorObj Out int Dropped int frames int delay Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acquireStamped Acquire an image from a frame grabber with current counter
263. es are used directly in the output image If method is set to ColorCorrect color correction using neighbouring pixel values is applied The color correction method usually produces better looking results when the image contains sharp vertical or horizontal edges Bayer encoded images normally start with a green pixel in the first column of the first row and a red pixel in the second column of the first row However if subimages are extracted from a Bayer encoded image the pattern may be offset The start parameter allows the operator to process such subimages When set to GR the operator treats the first and second pixels of the first row of the Bayer image as green and red respectively This setting works with standard Bayer images and subimages in which the upper left corner is located at an even row and even column When set to RG the two pixels are treated as red and green This setting is used with subimages in which the upper left corner is located at an even row and odd column When set to BG the two pixels are treated as blue and green This setting is used with subimages in which the upper left corner is located at an odd row and even column When set to GB the two pixels are treated as green and blue This setting is used with subimages in which the upper left corner is located at an odd row and odd column The start setting may be used for both encoding and decoding images Parameter Information direction Type int Default
264. es not allow the width or precision to be negative MSC sprintf unreliable e The n and p conversions are not supported Inappropriate in WiT Parameter Information format Type CorString 49 Demo iGraph Operators Blobs GetBlobFeatures igr Operators Blobs GetBlobFeaturesMasked igr Operators Blobs SelectFeatures igr C Prototype CorOpRtn cor_sprint CorObj Argl CorObj Arg2 CorObj Arg3 int Count CorString Outstr CorString format Headers include WITHOME h wConvert h Libraries WITHOME lib wObj lib WITHOME lib wConvert lib sscan Decode formatted input from a string in the style of C s sscanf i siah Description The sscan operator deals with formatted input in the style of C s sscanf function The operator breaks the input string down into its elements as controlled by the format parameter These elements are collected into a vector to form the output The format parameter is a sscanf style format string that contains the conversion specifications and directs the interpretation of the input The format parameter may contain one or more of the following e White space characters ie spaces tabs which are read and not stored One white 50 space character in the format string matches any number including zero of white space characters in the input string e Non white space characters except for the percent sign which are expected to match the
265. eses 265 WHT ON rreri EE EAA AE TEE ets pa tscale rrn prne EaR RAR E 266 Wr itelIMa pE e eaa E AEA E E patSc leRotaten inen a 267 writeObj searchEdit Transforms hiore e aae iaa esnie E Sedreh RUM se ere eea e e E O T EE E E SmatW ebrenn rna EE R A webClusterFeatures webGetBlobFeatsRoi WEDGELBIODF CAtur eS ccccccccccccccessccsesssesessseeessses 274 webGetBlobs StmartMatrix ccs sciences tighed ed cai eee 220 webGetBlobsRoi 278 AA 1O PEEN E E A 220 webGetPerimetet sssrinin 280 WEDLADEIBLIOD occecccccccccseccccesscsessssceestscsesssseseees 281 WEDLINCCOUNLETL ccccccseccsessscceessecceeseceesssseeeees 282 WEDPIOLFCALUIES cccccccccccccessccsesssceesssceeessseeeeens 283 webRealign D WEDROWEXIENA ccccccccessccsessssssesscesesscsessseeeeses 286 WEDROWREMOVE ccccccssssecesesssesesssceeensesesssseseeses 286 webScrollGraphics webWaterfall Frame Grabber and Hardware c ssssssssssssseee 289 SerlallOss 424 cesses end nnan evens openSerial CloseSeriahs via i ar aa A hese VEGUSCV IAL oes aici sieis axel R 290 writeSerial Video Acquisition cceeeeseeeeseeeeeteeeseeteeeaeeaees GCG COIOTREY aessa ances acqCounterRead acqCounterReset acqDisplay acqRgbBrightness sssr 299 ACQSCLUD i aiaia E E S 299 acqSize ioe acg SnapCOUNt sesinin iani 300 ACIN GUTE OS ina e E E 301 acqStart Sve ACG SUOD Sax fncoses E EE E E 302 ACQSWitCAGTAD ec
266. esh Type float Default 0 75 applyPatterns Type int Default 0 Values All 0 Sample 1 patSampling1 Type int Default 1 patSampling2 Type int Default 1 output Type int Default 0 Values Final Values 0 Candidates 1 maxMatches Type int Default 1 relThresh Type float Default 0 xVicinity Type int Default 0 yVicinity Type int Default 0 C Prototype CorOpRtn cor_patFind2D1 CorImage Target CorGraphic ROI CorPatModellVector PatModels CorFpointVector Locations CorFloatVector Scores CorIntVector Indexesl CorIntVector Indexes2 float rejectThresh float acceptThresh float minContrast int maxTilesOmitted int localSearch int subPixel int resolution int scale int maxCandidates float candidateThreshold float candidateRelThresh int applyPatterns int patSamplingl int patSampling2 int output int maxMatches float relThresh int xVicinity int yVicinity Headers include WITHOME h wSearch h Libraries WITHOME lib wObj lib WITHOME lib wSearch lib patinit1 Initialize pattern from image or image vector peiri IT o CT Description The patInit1l operator prepares a pattern model from an image or vector of images for use with the pattern matching operators patFind1 patLocall and patRoil The pattern model images may be broken up into equal sized rectangular subimages or tiles The pattern matching operators can optionally use a subset
267. esponding output vector element is the empty string Parameters noise The noise parameter specifies the minimum low to high or high to low change in pixel value necessary to constitute the edge of bar Parameter Information noise Type int Default 5 Demo iGraph Operators Matrix EAN 13 igr C Prototype CorOpRtn cor_getEAN13 CorImage Target CorObj ROI CorObj Decoded int noise Headers include WITHOME h wMatrix h Libraries WITHOME lib wObj lib WITHOME lib wMatrix lib getEAN8 Finds and decodes EAN 8 bar code WHI i Description The getEAN8 operator accepts a binary or grayscale image and one or more rectangular 225 regions of interest ROIs then searches the ROIs for EAN8 barcodes If a barcode is found in a region it is decoded to a string Inputs Target The input image must be an unsigned 8 bit grayscale image with the barcode represented as dark bars on a light background Roi A single ROI may be specified by a rectangular Graphic or Geometry object Multiple ROIs can be specified by a vector of Graphic or Geometry objects Outputs Decoded If the ROI is a single object then the output is a String object If a valid barcode is found the output string contains the characters represented by the barcode If no valid barcode is found an empty string is output If the ROT is a vector of rectangles then the operator seeks for a barcode in each ROI The output is a vecto
268. f pixels in the blob xmin Minimum x coordinate of blob left edge xmax Maximum x coordinate of blob right edge ymin Minimum y coordinate of blob top edge ymax Maximum y coordinate of blob bottom edge sumx Sum of all x coordinates sumy Sum of all y coordinates sumxsq_ Sum of all x coordinates squared sumysq Sum of all y coordinates squared sumxy Sum of all x coordinates y coordinates perimx X coordinate of a pixel on the blob perimeter lower right corner Y coordinate of a pixel on the blob perimeter lower right corner xc X coordinate of blob centroid sumx n ye Y coordinate of blob centroid sumy n angle Angle of best fit ellipse major axis clockwise from x axis major Length of best fit ellipse major axis minor Length of best fit ellipse minor axis axratio minor major xdiff xmax xmin 1 ydiff ymax ymin 1 perimy boxarea_ xdiff ydiff bxaratio npixels boxarea perim Perimeter data for blob must run the getPerimeter operator to fill in this structure Blobs The Blobs output is a vector of Blob objects Each Blob object contains a single run length The run lengths are grouped into contiguous regions Within each group the run lengths are sorted by increasing line number and then by the x location of the end of the run The groups are sorted on the last run length in the group The operator keeps a running line count the run lengths and info
269. f the file does not exist the output is 1 in the Integer format and an empty string in the String format Parameter Information filename Type CorFile Default sample request Type int Default 0 Values Exists 0 Size 1 Time 2 format Type int Default 1 Values String 0 Integer 1 C Prototype CorOpRtn cor_fileStatus CorObj Status CorFile filename int request int format Headers include WITHOME h wSystem h Libraries WITHOME lib wObj lib WITHOME lib wSystem lib getClipboard Get an image from the clipboard i H F Description The getClipboard operator reads the current contents of the system clipboard when it receives any object at its input If the current contents of the clipboard is a BMP format image getClipboard will read in the image and output it as a WiT color image If the clipboard contains anything other than a BMP format image an error occurs C Prototype CorOpRtn cor_getClipboard CorImage Out Headers include WITHOME h wSystem h Libraries WITHOME lib wObj lib WITHOME lib wSystem lib getTime Get the current date and time aoe Description The getTime operator reports the current date and time when any token is received at the input The format of the output is controlled by the format parameter The Integer format output is the number of seconds since a standard reference date The String format output is the date and local time
270. fast executin The output image type is the signed version of the input image type Parameter Information direction Type int Default 0 Values horizontal 0 vertical 1 optimization Type int Default 1 Values preciseness 0 speed 1 Demo iGraph Operators Filter Entropy Operators Filter Variance C Prototype CorOpRtn cor_prewitt CorImage In CorImage Out int direction int optimization CorOpRtn cor_prewitt_consume CorImage In int direction int optimization 90 Headers include WITHOME h wFilter h Libraries WITHOME lib wObj lib WITHOME lib wFilter lib refineEdges Finds edges on the image MB Of bone Description The refineEdges operator finds edges in images by detecting areas of high slope using Prewitt or Sobel gradient type operations Parameter threshold is applied to the image to eliminate weak edges The refineEdges operator uses derivatives calculated using either Prewitt or Sobel kernels as specified by the type parameter to approximate the gradient magnitude Only the maximum value of the gradient in the vertical or horisontal direction is left Parameter Information type Type int Default 0 Values Prewitt 0 Sobel 1 threshold Type float Demo iGraph Operators Filter Entropy Operators Filter V ariance C Prototype CorOpRtn cor_refineEdges CorImage In CorImage Out int type float threshold Headers include WITHOME h wFilter h
271. fault 1 290 C Prototype CorOpRtn cor_closeSerial CorObj In CorObj Out int port Headers include WITHOME h wSerial h Libraries WITHOME lib wObj lib WITHOME lib wSerial lib readSerial Read a string from a serial port t 3 mn Description The readSerial operator reads characters from a previously opened serial port when it receives any input object The type of output is specified by the output parameter When this parameter is set to String the operator produces a WiT string When it is set to CharVector a vector of characters is produced The vector of characters output may be necessary when a null character is expected in the data to avoid treating the null character as an end of string marker The port parameter specifies the serial port to be read The port parameter value 0 corresponds to the COM1 port parameter value 1 corresponds to COM2 and so on The port must have previously been opened using the openSerial operator Attempting to read a port that has not been opened results in an error The maximum number of characters to be read can be set using the maxChars parameter The operator will wait until the specified number of characters have been read or until a read timeout is reached Timeouts are set when the port is opened by the openSerial operator Parameter Information port Type int Default 1 maxChars Type int Default 1 output Type int Default 0 Values St
272. fault 100 reset Type int Default 0 Values No 0 Yes 1 Demo iGraph Operators Calculate Lut C Prototype CorOpRtn cor_randSequence CorVector Out CorUshortVector NextSeed int useSeed CorUshortVector seed int number int type float low float high int reset Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib 27 random Produce a random integer or float number I Description The random operator produces a random integer or float number from a uniform distribution over a specified range The type of the output value is specified by the type parameter and the range of numbers from which it is selected is specified by the low and high parameters The range for integers is ceil low floor high so if the low and high values are integers they are both included in the range The range for floats is low high that is the low value is included in the range the high value is not The random number sequence used to generate the numbers can be controlled using the useSeed seed and reset parameters A random number seed is generated the first time an operator executes or when the reset parameter is set to Yes starting a new random number sequence When the reset parameter is set to No the random number sequence is a continuation of the sequence used the last time the operator executed When the useSeed parameter is set to No the initial
273. ferences are used to produce a difference measure Root mean square RMS average squared error average absolute value of the error euclidean distance sum of squared errors sum of absolute value of errors or maximum error may be selected Input images may be any of the standard WiT image types unsigned 8 bit signed 8 bit unsigned 16 bit signed 16 bit float complex RGB HSV YUV The difference measure is calculated over all pixels in the intersection of the input image ROIs If the useMask parameter is set to Yes a mask specified by the mask parameter is applied so that only pixels in the mask region and in the intersection of the ROIs of teh two input images are used to calculate the difference Input vector elements may be any of the types valid for images or unsigned 32 bit integer signed 32 bit integer or double The two input vectors must be the same size Parameter Information measure Type int Default 3 Values RMS 0 Mean Square Error 1 Average Difference 2 Euclidean Distance 3 Sum of Squared Errors 4 Sum of Differences 5 Maximum 6 useMask Type int 199 Default 0 Values No 0 Yes 1 Mask Type CorObj Default OBJ_B T CorGeom 4 CorFpoint 2 0 0 99 99 Demo iGraph Operators Filters V ariance Operators Filters Entropy C Prototype CorOpRtn cor_difference CorObj In0 CorObj Ini float Difference int Number int measure int useMask CorObj Mask Headers includ
274. ffer Description The buf operator passes its input object to its output This operator is used primarily as a means to connect the output of an operator back to its input Connecting outputs directly to the inputs of the same operator causes deadlock This form of deadlock is corrected by using the buf operator between the output and input connections cCode Dynamically compiled C code 4 Description Dynamically compiled C code clearObjs clear data objects remaining in an igraph a a ny 112 Description Igraphs that have loops and branches and operators with multiple inputs may have extra data objects remaining from previous iterations which can block proper execution of subsequent iterations This is especially common when error conditions are encountered during execution clearObjs can be used to clear all remaining data objects in an igraph The output is the number of data objects freed Parameter Information applyToLevels Type int Default 0 Values current 0 current and lower 1 lower only 2 Demo iGraph Operators clearObjs collector Collect input objects and output a vector n it Description The collector operator collects a set of input objects of the same type and forms a vector as output The objects which form the vector are collected from the top input port The Control input port expects an integer object representing the collector state The number of objects collecte
275. fies the width of the strokes or dots that make up the characters in the font outType The outType parameter is used to specify whether the images in the output font are gray scale or binary Gray scale images may only be used with used with the ocrGrayInit operator Binary images may be used with either the ocrGrayInit or octPerimlInit thresh The thresh parameter is used to specify a threshold value for binary font images It is given as a percent of the full scale value invert If the invert parameter is set to No gray scale characters are light on a dark background If it is set to Yes gray scale characters are dark on a light background filter This parameter allows gaussian smoothing to be applied to the font images The user can specify no smoothing or a gaussian filter of up to 7x7 pixels to be applied to the font images For binary images the smoothing is applied before thresholding charSet When the charSet parameter is set to All the output font includes the digits upper case characters and the characters and When it is set to Alpha numeric it contains the digits and upper case letters When it is set to Digits it contains only the digits Parameter Information style Type int Default 1 Values Point 0 Line 1 width Type float Default 20 height Type float Default 40 clearZone Type float Default 10 strokeWidth Type float Default 5 outType Type int Default 1 Values binary 0
276. fitCircleToPts CorObj Points CorGraphic Circle int Iterations int autoInit CorFpoint initCenter int maxIterations Headers include WITHOME h wFit h 102 Libraries WITHOME lib wObj lib WITHOME lib wFit lib fitLineToPts Least squares line fit to points Description The fitLineToPts operator performs a least squares fit to fit a line to the input points The operator s input may be a vector of Point Fpoint PtValue Graphic Geometry or Edge objects or a single Graphic or Geometry object If the input is a vector of Graphic Geom or Edge object each object in the vector must of point type If the input is a Graphic or Geometry object it may be any type except circle and text In this case the line will be fit to the vertices treated as independent points If the input is a vector of PtValue objects the value field of the elements of the vector will be used as a weight value for a weighted least squares fit Negative weight values are ignored The operator s output is a single line Graphic object The endpoints of the line are the extremes of projected positions of the points in the input set onto the best fit line The direction of the line follows the relative position of points generating the endpoints in the input Demo iGraph Operators Geometry convexHull igr C Prototype CorOpRtn cor_fitLineToPts CorObj In CorGraphic Out Headers include WITHOME h wFit h Libraries
277. g Default mem Demo iGraph Operators clearObjs memStore Store objects for later recall Description Store an object in memory so that it can be recalled later anywhere in an igraph as many times as required Objects stored in memory are referenced by the name parameter which refers to a global name See chapter Operators Special in User s Manual for details See also memFree memRecall Parameter Information name Type CorString Default mem Type CorString The global name of the memory location persist Type int Default 0 Type choice no 0 yes 1 If persist is no the memory will be freed when a new igraph is loaded Otherwise the memory will be valid even when a new igraph is loaded This is useful when a large object needs to be read from disk and can be processed by multiple igraphs Values no 0 yes 1 Demo iGraph Operators clearObjs mux5 Multiplex input to 1 of 5 possible outputs Shiki Description Multiplex input to 1 of 5 possible outputs Parameter Information choice Type int Default 0 Values 0 0 1 1 2 2 3 3 4 4 Demo iGraph Operators clearObjs mux9 Multiplex input to 1 of 9 possible outputs bed i T POET Description Multiplex input to 1 of 9 possible outputs Parameter Information choice Type int Default 0 Values 0 0 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 Demo iGraph Operators clearObjs oneshot
278. g y axis from y centroid at max radial a maxLength Max Max length of perimeter of perimeter minLength Min length of perimeter mawaa Max width of a minWidth Min width of Min width of perimeter points Point vector containing all perimeter points The operator also outputs the blobs in the form of a Blob vector The Blobs output is a vector that contains the run lengths that make up the blobs in the image Each element of the vector consists of an id the row start and end of the run length and the label value of the run length Run lengths belonging to the same connected object in the image have the same id number which also is the id number of the corresponding element of the BlobFeatures vector The type parameter is used to specify whether the input image is interpreted as a binary or labeled image In a binary image all non zero valued pixels are equivalent and blobs are connected regions of non zero pixels Binary images are typically produced by the thresh operator In a labeled image equal valued pixels are equivalent In this case the blobs are connected regions of equal valued pixels Labeled images are produced by operators such as kMeans and watershed The connected parameter is used to specify whether the regions are edge connected 4 or corner connected 8 The perimeters of the blobs will be traced and included in the Blob features if the perimeter parameter is set to Yes T
279. graph based on a comparison between two values If the condition parameter is set to Empty then the lower input can be a vector image or string If the vector or image is empty or if the string is NULL or empty then the condition will be true For all other values for condition the lower input should be a numeric object such as an integer or float ifConditional compares it to the value set in the constant parameter The type of comparison is chosen using the condition parameter It may be one of gt lt gt or lt If the condition is true the object received at the top input port is passed to the top branch labelled with TrueCond Otherwise the input object is passed to the bottom branch labelled with FalseCond See also ifCond2 ifCond3 Parameter Information condition Type int Default 0 Values 0 gt 1 lt 2 3 gt 4 lt 5 Empty 6 constant Type CorString Default Demo iGraph Operators clearObjs in Hierarchical operator input port in In Object 2H Out Object Description The in operator is used to associate an input port of a super node to an entry point in its sub graph The instance name of in is used as the name of the port in the super node Therefore you should choose meaningful names for each instance of the in operator Demo iGraph Operators clearObjs label Goto operator destination Description The label operator serves as a des
280. graphicType int linePen int lineR int lineG int lineB int pointPen int pointkR int pointG int pointB Headers include WITHOME h wConvert h Libraries WITHOME lib wObj lib WITHOME lib wConvert lib scanString Convert a string to a vector of substrings des ok Description The scanString operator allows the user to extract substrings from the input string and assemble these substrings into a vector of strings The input string is divided at any character or combination of characters from the string entered in the splitChars parameter The characters from splitChars string are not include in the substring For example the input string 1 2 3 scanned with the splitChars string where the last character in this string is a space will result in the a three element vector containing the strings J 2 and 3 Multi column vectors of strings are no longer supported The nCols parameter has no effect all settings are equivalent to a setting of 1 in versions prior to WiT 6 0 47 Parameter Information splitChars Type CorString Default nCols Type int Default 1 Demo iGraph Operators Blobs GetBlobFeatures igr Operators Blobs GetBlobFeaturesMasked igr Operators Blobs SelectFeatures igr C Prototype CorOpRtn cor_scanString CorString In CorStringVector Out CorString splitChars int nCols Headers include WITHOME h wConvert h Libraries WITHOME lib w
281. h Headers include WITHOME h wConvert h Libraries WITHOME lib wObj lib WITHOME lib wConvert lib vectorsTolmage Convert a vector of vectors into an image zo Description The vectorsToImage operator creates an image from a vector of vectors with the element vectors the vectors that are elements of the input vector making up the rows of the image The base type of the element vectors can be any type including 51 vectors images or user defined types The output image will be the same type as the element vectors WiT processing operators will generally only accept images from among the scalar types uint8 int8 uintl6 int 16 and float32 color types RGB HSV and YUV and the Complex type In addition the display operators will display these types as rendered images all other types are displayed in a text format similar to the vector text display format The element vectors may be of varying length The output image width will be equal to the length of the longest element vector Empty image elements pixels will be filled with zero valued versions of the element vector base type This may result in inconsistent data values for compound data types such as the Graphic type The setRoiWidth parameter allows the user to specify whether the the sith of the intrinsic image ROI should be set to the width of the shortest element vector by selecting Min or to the length of the longest element vector the entire i
282. h ch2 Low and ch2 High parameters which give the minimum and maximum extent of the volume along the channel 0 channel 1 and channel 2 axes If the input image is an HSV image the channel 1 volume wraps around at 255 so if ch0 High is less than ch0 Low the volume extends from ch0 Low to 255 and from 0 to ch0 High The invert parameter controls the value assigned to pixels in the output that correspond to input pixels with values in the color space volume When the parameter is set to No these pixels are set to 255 and the remaining pixels are set to 0 When it is set to Yes these pixels are set to 0 and the remaining pixels are set to 255 Parameter Information type Type int Default 0 Values Level 0 Box 1 Ellipsoid 2 Cylinder 3 ch0 Low Type int Default 0 Values 0 255 cho Type int High Default 255 Values 0 255 chl Low Type int Default 0 Values 0 255 chl Type int High Default 255 Values 0 255 ch2 Low Type int Default 0 Values 0 255 ch2 Type int High Default 255 Values 0 255 invert Type int Default 0 Values No 0 Yes 1 Demo iGraph Operators Segmentation ColorThresh Operators Segmentation ColorThreshHS V C Prototype CorOpRtn cor_colorThresh CorImage In CorImage Out int type int ch0O_Low int ch0_High int chi_Low int chl_High int ch2_Low int ch2_High int invert Headers include WITHOME h wSegment h Libraries W
283. h a histogram oh Description The smoothHist operator smoothes a histogram produced by the histogram or histRoi operator The output histogram frequency values are 203 smoothed by averaging the frequency values over a neighbourhood determined using width parameter The width is specified in terms of the image pixel values the x coordinates of the histogram entries rather than number of bins The width value is converted to the nearest odd number of bins based on the difference between the pixel values associated with the first two entries in the histogram Consequently the entries in the input histogram must be equally spaced in the pixel value x coordinate as they are when produced by the histogram and histRoi operators Parameter Information width Type float Default 3 Demo iGraph Operators Statistics HistThresh C Prototype CorOpRtn cor_smoothHist CorFpointVector Hist CorFpointVector SmoothedHist float width Headers include WITHOME h wStats h Libraries WITHOME lib wObj lib WITHOME lib wStats lib stats Compute basic image statistics hi td zoa o err Description The stats operator derives basic statistics for an image or vector input The operator accepts integer 204 or float images and vectors of integer unsigned integer short unsigned short float or double type values A geometric or image mask can be specified for image inputs The number of value
284. he labels scores and locations of the candidates rather than those of the refined matches normally produced The Candidates setting is usually used during development to help determine the best settings for the various parameters that control candidate selection Parameter Information maxMatches Type int Default 1 xVicinity Type int Default 0 yVicinity Type int Default 0 rejectThresh Type float Default 0 acceptThresh Type float Default 1 relThresh Type float Default 0 minContrast Type float Default 0 1 scale Type int Default 1 Values 2 0 4 1 8 2 16 3 32 4 maxCandidates Type int Default 400 candsInVicinity Type int Default 4 candThreshold Type float Default 0 5 candRelThresh Type float output Default 0 5 Type int Default 0 237 Values Final Values 0 Candidates 1 C Prototype CorOpRtn cor_ocrGrayRun_1 CorImage Target CorObj ROI CorFontModelGrayVector FontModel CorVector Labels CorVector Scores CorVector Locations int maxMatches int xVicinity int yVicinity float rejectThresh float acceptThresh float relThresh float minContrast int scale int maxCandidates int candsInVicinity float candThreshold float candRelThresh int output Headers include WITHOME h wOcr h Libraries WITHOME lib wObj lib WITHOME lib wOcr lib ocrLocalBin Binarization using local thresholding we Description The ocrLocalB
285. he input index vector There does not have to be a one to one or onto relationship between the index and object vectors The operator may be used to extract a subset of the object vector and index values may be repeated Index values less than zero or greater than maximum object vector index result in an error The shuffleVector operator is most often used with the sort or partition with their output parameter set to index In this mode these operators produce a vector of index values corresponding to the sorted object vector or the selected subset respectively The index values can then be used to sort or partition a vector related to the vector on which the operation was performed C Prototype CorOpRtn cor_shuffleVector CorVector In CorIntVector Indexes CorVector Out Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib sortObj Sort a vector of objects on a named field aE Description The sortObj operator allows the user to sort a vector of simple numeric objects or complex objects Sorting of complex objects is based on the value of a particular field within the object The field must be one of the numeric or character basic types The field to be sorted on is specified using the field parameter If the elements of the input vector are simple numeric types the the field parameter must be left empty The vector can be sorted in ascending or descending
286. he parameters on the Select subpanel allow the operator to ignore blobs that are smaller or larger than specified sizes The minPixels minWidth and minHeight parameters specify the minimum number of pixels the minimum width extent in the x direction and the minimum height extent in the y direction respectively of blobs produced The maxPixels max Width and maxHeight parameters specify the maximums for the corresponding features If a maximum parameter value is set to zero no maximum limit is applied for the corresponding feature If the useMask parameter is set to Yes the blobs use only pixels in the intersection of the image ROI and a mask specified by the Mask parameter The mask can be specified by a Geometry or Graphic object or by a binary image The mask may also be specified by a vector of Geometry or Graphic objects binary images or Blobs In this case pixels may be used more than once if the elements overlap in the image Parameter Information type connected perimeters useMask mask minPixels minWidth minHeight maxPixels maxWidth Type int Default 0 Values binary 0 labeled 1 Type int Default 1 Values 4 0 8 1 Type int Default 0 Values No 0 Yes 1 Type int Default 0 Values No 0 Yes 1 Type CorObj Default OBJ_B T CorGeom 4 CorFpoint 2 0 0 99 99 Type int Default 0 Type int Default 0 Type int Default 0 Type int Default 0 Type int Default
287. hich acquire is called repeatedly to deliver single frames for processing Operators that configure the acquisition for example acqSize pproCustom should not be called after acqStart Continuous transfer is stopped automatically when an igraph stops Alternatively the transfers can be stopped explicitly using the acqStop operator Parameter Information queueSize Type int Default 2 C Prototype CorOpRtn acqStart int queueSize Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqStop Stop ping pong acquisition on frame grabber a a Description The acqStop operator halts continuous mode image transfer for a frame grabber See the help for the acqStart operator for more information on continuous mode acquisition C Prototype CorOpRtn acqStop 302 Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqSwitchGrab Grab up to 4 cameras setup for switching a k E G Description The acqSwitchGrab operator grabs images from several cameras on the currently installed frame grabber Typically the Frame Grabber tool in WiT is used to configure the frame grabber for switched acquisition using 2 3 or 4 cameras that are constantly being multiplexed between frames or fields All cameras must be genlocked with the same synchronization to ensure switching is reliable This operator supports up to 4
288. histogram corresponds to the average of the pixel values that formed the image When the centroid is calculated for numeric vectors the element values correspond to frequency and the vector index corresponds to value Demo iGraph Operators Statistics LocateInProj_Centroid C Prototype CorOpRtn cor_histCentroid CorVector Hist float Centroid Headers include WITHOME h wStats h Libraries WITHOME lib wObj lib WITHOME lib wStats lib histPeaks find histogram maxima or minima LL Description The histPeaks operator finds local maxima or minima in a histogram produced by the histogram or histRoi operator The operator output is a vector of histogram entries Real point objects in which the x coordinate is the pixel value and the y coordinate is the frequency value associated with the maximum or minimum The type parameter is used to specify whether the operator will find minima or maxima in the histogram The vicinity parameter is used to suppress multiple peaks in close proximity If a higher peak is found in the vicinity of a local peak the lower peak is suppressed The higher peak may be suppressed by an even higher peak in its vicinity which may not be in the vicinity of the lowest peak In this case only the highest peak is retained even though the lowest peak is outside of its vicinity The vicinity is expressed in pixel value units corresponding to the x coordinate of the histogram entries
289. i CorImage Out int channel Headers include WITHOME h wColor h Libraries WITHOME lib wObj lib WITHOME lib wColor lib splitChannels Split multichannel image into its three component channels aast Description The splitChannels operator separates a multichannel input image into three single channel output images The input image must be a RGB HSV or Yuv image The output images are 8 bit unsigned images For RGB HSV and Yuv images respectively ch0 is the red hue or Y intensity channel ch is the green saturation or u channel and ch2 is the blue value or v channel Demo iGraph Operators Color HS V C Prototype CorOpRtn cor_splitChannels CorImage In CorImage Ch0O CorImage Chi CorImage Ch2 Headers include WITHOME h wColor h 38 Libraries WITHOME lib wObj lib WITHOME lib wColor lib Complex The Complex library provides support for complex images The library includes operators that create complex images from real images and operators that extract real components or derived images from complex images complexMerge Merge real imaginary component images into a complex image Description The complexMerge operator converts two input images into a complex output image The input parameter determines whether the input images are interpreted as the the real and imaginary components of the complex image or as the magnitude and phase The operator
290. i such that image I n is the result of applying n reduce operations to the input image All other images I i where i ranges between 1 and n are differences between the input image after i 7 reduce operations and the expanded version of the input image after i reduce operations Notationally Let G 1 be the input image Let G i denote an image sequence where G i reduce G i 1 Then I i G i expand reduce G i Unless i n then I n G n The filter parameter selects the type of gaussian filter to be applied The effect of each filter is described in the WiT filter library during the discussion of the gauss operator The minsize parameter defines the minimum size that an image G i can be before a reduce operation will not be performed and pyramid generation will terminate Parameter Information filter Type int Default 0 Values None 0 2x2 1 3x3 2 5x5 3 7x7 4 minsize Type int Default 8 C Prototype CorOpRtn cor_burt CorImage In CorImageVector Pyramid int filter int minsize Headers include WITHOME h wPyramid h Libraries WITHOME lib wObj lib WITHOME lib wPyramid lib burtRecon Reconstruct original image from a Burt pyramid 183 fe patya Description The burtRecon operator reconstructs an original image from the Burt pyramid that was generated from it This reconstruction should be exact barring things like overflow during computation Given
291. ical objects The forward matrix output at the FwdXform port specifies the transformation that takes the points input at the From lower input port to the points input at the To upper port The reverse matrix output at the RvsXform port is its inverse The input points may be specified by vectors of Point Fpoint or point type Graphic or Geometry objects The inputs must contain the same number of points There must be four or more points in the input vectors If more than four points are specified the transformation is determined using a least square fit The two sets of input points are matched pairwise in the order that they appear in the matrix Alternately the upper input points may be specified by an image In this case the points are the corners of the image in clockwise order from the upper left If the upper input is an image the lower input vector must contain four points The getWarpTransform operator is usually used to generate matrices for use with the warp operator The matrices may also be used with the warpFeatures and warpPerimeter operators Demo iGraph Operators Fit warp igr C Prototype CorOpRtn cor_getWarpTransform CorObj To CorObj From CorFloatVector FwdXform CorFloatVector RvsXform Headers include WITHOME h wGeom h Libraries WITHOME lib wObj lib WITHOME lib wGeom lib 125 insertGraphic Move graphics to a new image position Description The insertGraphic
292. ideo When the enable parameter is turned on live video will be presented in a window for display The live window will be closed automatically when an igraph stops execution or when this operator is run again with the enable parameter set off Parameter Information enable Type int Default 0 Values no 0 yes 1 Demo iGraph Operators Video Acquisition trigger C Prototype CorOpRtn acqLive int enable Headers include WITHOME h wVideo h 297 Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqLut Load an input LUT on frame grabber Hh Description The acqLut operator is used to load a hardware lookup table which may be available on the frame grabber The input is a Vector of integers which holds the values of the lookup table to be loaded The lutGen operator can be used to generate many common lookup tables Parameter Information lutNumber Type int Default 0 lut Type CorVector Default 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121
293. if the conditional is true and 0 if false 18 Parameter Information condition Type int Default 0 Values 0 gt 1 lt 2 3 gt 4 lt 5 Demo iGraph Operators Transforms FourierFilter C Prototype CorOpRtn cor_conditional CorObj In0 CorObj In1 int Out int condition Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib div Divide one object with another Description Divide the first input by the second See add for more information about supported types Dividing a scalar first input by a vector or image is not allowed Demo iGraph Operators Transforms FourierFilter C Prototype CorOpRtn cor_div CorObj A CorObj B CorObj quotient Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib equalize Linear image rescaling a Sale r Description The equalize operator may be used to re map the frequency distribution of an image Pixel values between the range fromLo to fromHi can be compressed or expanded to the new range toLo to toHi The contrast of an image for example may be increased by choosing a from range which represents a greyscale area of interest and mapping these pixels to cover the entire range of the image Parameter Information fromLo Type float Default 0 fromHi Type float Default 255 toLo Type float Default 0 toHi
294. igid body transformation matrix for alignment 124 Skee ie al jolie Description The getRigidBodyTransform operator creates two 3x3 matrices for forward and inverse rigid body transformations that produce the best fit between the reference points input at the lower input port and the test points input at the upper port The operator can be made to assume a one to one correspondence between the input point vectors or can search for the best fit between the two point sets The input points may be specified by vectors of Point Fpoint or point type Graphic or Geometry objects If the correspondence parameter is set to 1 1 both input vectors must contain exactly three points This mode is used when the points in the test vector are known to be in the same order as the points in the reference vector If the correspondence parameter is set to Best the reference point vector must contain three points and the test point vector must contain three or more points The operator will search for the permutation of three points from the test vector that gives the best fit to the reference points under a rigid body transformation In the Best correspondence mode the user can select minimum and maximum rotation angles using the minAngle and maxAngle parameters These parameters specify the limits on the angle by which the reference points are rotated before the fit to the test points is evaluated The output matrices can be used to transform im
295. il the graph stops Blobs which reach the bottom of the last image processed will not be output by the operator resetBlobId The blob id number can be set to zero with the resetBlobId parameter This will set the next blob id to zero after processing the current image This parameter is usually set to No or used as an input parameter It may be tied to the reset parameter when both are input parameters in which case a new set of id numbers will be started every time the operator is reset Parameter Information type Type int Default 0 Values binary 0 labeled 1 connected Type int Default 1 Values 4 0 8 1 maxBlobHeight Type int reset resetBlobld minPixels minWidth 276 Default 0 Type int Default 0 Values No 0 Yes 1 Type int Default 0 Values No 0 Yes 1 Type int Default 0 Type int Default 0 minHeight Type int Default 0 C Prototype CorOpRtn cor_webGetBlobFeatures CorImage In CorFeatureVector BlobFeatures CorBlobVector Blobs int type int connected int maxBlobHeight int reset int resetBlobld int minPixels int minWidth int minHeight Headers include WITHOME h wWeb h Libraries WITHOME lib wObj lib WITHOME lib wWeb lib webGetBlobs Run length encode blobs in a stream of line scan images ft Description The webGetBlobs operator performs pixel connectivity analysis on a stream of binary or labeled images The operator ext
296. image in which case the parameter image s ROI is used The graphic geometry and image type parameters are typically input parameters produced by other WiT operators The ROI specified by the parameter will be clipped to fit the input image if it extends outside the image Parameter Information ROI Type CorObj Default OBJ_B T CorVector CorFpoint 2 0 0 255 255 C Prototype CorOpRtn cor_setRoi CorImage i0 CorImage o0 CorObj ROI CorOpRtn cor_setRoi_consume CorImage i0 CorObj ROI Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib shuffleVecExpr Reorder a vector using a C code like expression Description The shuffleVecExpr operator selects and rearranges the elements of an input vector based on a mapping specified by a C code like expression The input vector elements may be of any type The Out output vector is made up of elements from the input object vector specified the expression The Indexes output vector is an integer vector that gives the mapping from the input vector to the output vector The mapping from the input to output is specified by the expression parameter This parameter is a string that uses a C code like syntax to specify input vector indexes as a function of the output vector indexes The symbol A is used to represent the output vector index in the expression The function can be parameterized by a user supplied constan
297. image the parts of the model image that fall outside the input image are assigned a background value calculated by averaging the comer pixels in the overlapping part of the ROI The label value of the Graphic object is used as the the label associated with the model If the the label field of the graphic is empty an empty string is used as the label for the corresponding model image If the ROI is specified with Geom objects its index number in the input vector is used as a label 0 if a single Geom Duplicate labels are allowed If the Font input is an OcrFont type object the new model images and their corresponding labels are added to the font In all other cases a new font is created The characters model images and labels in the font will be sorted alphabetically by label if the sortByLabel parameter is set to Yes If it is set to No new characters will be appended to the characters in the existing font if any in the order of the corresponding ROIs in the input vector The operator s output is an OcrFont type object that can be used by a ocrGrayInit or ocrPerimInit operator to create a font model for use with the ocrGrayRun or ocrPerimRun operator The ocrGrayRun operator will run faster if all of the model images are the same size so when preparing a gray scale font the width and height parameters may be set to non zero values which will force the images to be the specified size When preparing a font for the ocrPerimIni
298. imeter operator computes perimeters of binary images and warps them into a perspective warped image space The warped space is specified by a transformation matrix input at the Transform input port of the operator The transformation matrix is typically produced by the getWarpTransform operator The perimeter is determined for each element in the Blob feature vector input at the Features input port by tracing the outline of objects in the unwarped binary image input at the SourceImage port As with the getPerimeter operator the binary image should be the image from which the blobs used to produce the Blob feature vector were extracted The connected parameter should be set to the same value as the connected parameter of the getBlobFeatures operator that produced the blobs Notice that in this case the warped perimeter will be correctly determined but it will not match the other features in the Blob feature object The perimeters produced by applying the getWarpedPerimeter operator to a Blob feature vector generated from the unwarped image are approximately equal to those produced by first warping the image then extracting the Blob feature vectors and applying the getPerimeter operator Parameter Information connected Type int Default 1 Values 4 0 8 1 type Type int Default 0 Values binary 0 labeled 1 Demo iGraph Operators Blobs WarpFeatures C Prototype CorOpRtn cor_getWarpedPerimeter CorFeatureVector Feat
299. in operator binarizes a text image based a threshold calculated for each pixel from the pixel values in its neighbourhood Inputs In The input image may be any integer image type or a float image 238 Outputs Out The operator produces a binary unsigned 8 bit image with white text on a black background Parameters background The background parameter is set to light if the input image has a light background with dark text or to dark if the input image has a dark background with light text Image pixels with a value below the threshold value are set true non zero and those above the threshold as set false zero if the background has been specified as light The opposite is true if the background has been specified as dark width and height The width and height parameters control the size of the size of the neighbourhood used to calculate the threshold for each pixel value threshFactor The threshold at each pixel is calculated based on the threshFactor parameter and the mean maximum and minimum pixel value in the neighbourhood of the pixel The threshFactor parameter is usually set to a value between 0 and 1 A value of 1 will give the local mean as the threshold value Lower values produce thresholds nearer the foreground value and usually result in thinning of the blobs in the binarized image Parameter Information background Type int Default 0 Values light 0 dark 1 width Type int Default 25 height Typ
300. in pixels The rest of parameters is the same as in the rotate operator Several of the parameters are used only when the input object is an image Parameter Information angle Type float Default 0 shift Type CorPoint sizeOp Type int Default 0 Values expand 0 clip 1 rotateAbout Type int 127 Default 0 Values center 0 user 1 rotatePt Type CorPoint Default 0 0 antiAlias Type int Default 1 Values no 0 yes 1 background Type int Default 0 Values auto 0 user 1 bgValue Type float Default 0 Demo iGraph Operators Fit warp igr C Prototype CorOpRtn cor_move CorObj In CorObj Out float angle CorPoint shift int sizeOp int rotateAbout CorPoint rotatePt int antiAlias int background float bgValue Headers include WITHOME h wGeom h Libraries WITHOME lib wObj lib WITHOME lib wGeom lib pan Pan an object 128 Description The pan operator shifts the input object along the x axis by the given amount The amount paramater controls the number of pixels the object is shifted A positive pan amount shifts the object to the right Negative amounts shift the object to the left In the case of the image pixels shifted off the right or left edges wrap around to the left or right sides respectively Parameter Information amount Type int Default 10 Demo iGraph Operators Fit warp igr C Prototype CorOpRtn cor_pan CorObj In CorObj
301. in the format Wed Nov 10 14 27 42 1999 Parameter Information format Type int Default 0 Values String 0 Integer 1 C Prototype CorOpRtn cor_getTime CorObj Time int format Headers include WITHOME h wSystem h Libraries WITHOME lib wObj lib WITHOME lib wSystem lib mkdir Create directory 207 Description Create directory Parameter Information dirname Type CorFile force Type int Default 0 Values yes 0 no 1 C Prototype CorOpRtn cor_mkdir CorFile dirname int force Headers include WITHOME h wSystem h Libraries WITHOME lib wObj lib WITHOME lib wSystem lib playSound Play a waveform sound file TE Description Play a waveform sound file Parameter Information filename Type CorFile Default type Type int Default 0 Values Asynchronously 0 Synchronously 1 loop Type int Default 0 Values off 0 on 1 208 nostop Type int Default 0 Values yes 0 no 1 C Prototype CorOpRtn cor_playSound CorFile filename int type int loop int nostop Headers include WITHOME h wSystem h Libraries WITHOME lib wObj lib WITHOME lib wSystem lib print Print an object Description The print operator converts the input object to printer compatible format or text and then either sends it to the printer or saves it to a file image quality If high grayscale images are output in 256 8 bit grayscale C
302. inconsistency in parameter naming is required for back compatability The output image is an 8 bit unsigned integer image in the grayscale mode and an RGB color image in the color mode In the color mode the background image produced when the overlay parameter is turned on is produced by setting the red green and blue channels for each pixel to the scaled input image value It may be necessary to select the Private colormap from the image properties panel to best display the color images In the grayscale mode this operator is useful if the ROI set is to be used extensively since the rasterization step is done only once and eliminates on the fly ROI computation The color mode is primarily for preparation of images for display Parameter Information overlay Type int Default 0 Values off 0 on 1 rangeType Type int Default 0 Values Actual 0 Max 1 User 2 min Type float Default 0 max Type float Default 255 outType Type int Default 0 Values grayscale 0 color 1 useColor Type int Default 1 Values original 0 force 1 outline Type int Default 0 Values off 0 on 1 penval Type int Default 255 Values 0 255 penR Type int Default 255 Values 0 255 penG Type int Default 255 Values 0 255 penB Type int Default 0 Values 0 255 fill Type int Default 1 Values off 0 on 1 pixval Type int Default 255 Values 0 255 fillR Type int Defaul
303. indicates no correlation between the pattern and target and one indicates a perfect match If no part of the target image matches the pattern sufficiently well empty output vectors are produced acceptThresh The acceptThresh parameter specifies an acceptance threshold above which a score value is always considered a valid match If the maximum number of matches specified by the maxMatches parameter are found with match scores above the acceptThresh value the search stops immediately This parameter is used to control early termination of the search for improved speed minContrast The minContrast parameter is used to specify the minimum contrast required in the target image to produce a match The parameter specifies the minimum standard deviation of the pixel values in a possible match region relative to the standard deviation of pixel values in the pattern model image This parameter is usually set to a value between zero and one although values greater than one can be used if the target image has higher contrast than the model images maxTilesOmitted If the pattern model breaks the pattern images into multiple tiles the maxTilesOmitted parameter allows the operator to omit some of the tiles while calculating the best matching score The operator will use the best score for a location that omits at most the number of tiles specified by this parameter The operator will always use at least one tile to calculate the best score
304. ine Parameter Information x Type int Default 0 y Type int Default 0 Demo iGraph Operators Video Acquisition trigger C Prototype CorOpRtn acqOverlay CorObj In int x int y Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqRgbBrightness Set video offset and gain for RGB frame grabber Description The acqRgbBrightness operator sets the contrast gain and brightness offset of the red green and blue bands of a color image acquired with the acquire operator The parameters are given as a value between 0 and 100 where 0 is the minimum and 100 is the maximum value available on the frame grabber hardware in use Parameter Information rBrightness Type int Default 50 Values 0 100 gBrightness Type int Default 50 Values 0 100 bBrightness Type int Default 50 Values 0 100 rContrast Type int Default 50 Values 0 100 gContrast Type int Default 50 Values 0 100 bContrast Type int Default 50 Values 0 100 Demo iGraph Operators Video Acquisition trigger C Prototype CorOpRtn acqRgbBrightness int rBrightness int gBrightness int bBrightness int rContrast int gContrast int bContrast Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqSetup Setup hardware dependent settings 299 Description The acqSetup operato
305. ing Parameters checkSum The checkSum parameter defines whether the bar code contains the check sum character noise The noise parameter specifies the minimum low to high or high to low change in pixel value necessary to constitute the edge of bar start and stop The start and stop parameters define whether the bar code contains the start and stop characters Parameter Information noise Type int Default 5 checkSum Type int Default 1 Values No 0 Yes 1 start Type int Default 1 Values No 0 Yes 1 stop Type int Default 1 Values No 0 Yes 1 Demo iGraph Operators Matrix BC412 igr C Prototype CorOpRtn cor_getBC412 CorImage Target CorObj ROI CorObj Decoded int noise int checkSum int start int stop Headers include WITHOME h wMatrix h Libraries WITHOME lib wObj lib WITHOME lib wMatrix lib getBarcode Finds and decodes any bar code Description Finds and decodes any bar code Parameter Information minSize Type int Default 20 maxSize Type int Default 40 invert Type int Default 1 Values No 0 Yes 1 noiseLev Type int Default 20 threshold Type int Default 10 barcodes Type int Default 254 Values bitwise OR 412 1 128 2 39 4 UPCa 8 ITF 16 EAN8 32 EAN13 64 UPCe 128 checkSum Type int Default 1 Values No 0 Yes 1 timeout Type int Default 100 Demo iGraph Operators Matrix BC412 igr C Pr
306. ing cannot be converted to the specified numeric type an error occurs halting the igraph Parameter Information type Type int Default 0 Values byte 0 ubyte 1 short 2 ushort 3 int 4 uint 5 float 6 double 7 String 8 File 9 Point 10 Real point 11 char 12 hex8 13 hex16 14 hex 15 conversion Type int Default 0 Values round 0 floor 1 ceiling 2 C Prototype CorOpRtn cor_cast CorObj In CorObj Out int type int conversion Headers include WITHOME h wConvert h Libraries WITHOME lib wObj lib WITHOME lib wConvert lib cvtCase Convert a string to all upper or all lower case dijm E Description The cvtCase operator converts the input character string src to either all upper case or all lower case depending on the to parameter Parameter Information to Type int Default 0 Values Upper 0 Lower 1 C Prototype CorOpRtn cor_cvtCase CorString Src CorString Dst int to Headers include WITHOME h wConvert h Libraries WITHOME lib wObj lib WITHOME lib wConvert lib formatString Concatenate a vector of strings into a string one abe E Description The formatString operator concatenates strings from the input vector and delimiter strings into a single string The delimiter strings are specified in the operator s parameters The colDelimit string is inserted after each element except the last The prepend string is inserte
307. ing edges produce negative values This operator could be executed in two modes The user may choose the mode by using the optimization parameter If the parameter value is speed then execution goes way faster otherwise operator delivers more precise results The operator accepts integer float type input images and color images in the case optimization paramiter is speed An unsigned integer input image produces a signed integer output image with the same number of bits as the input image In all other cases the output image is the same type as the input image The output is scaled so that for integer images a full scale step edge in the specified direction will produce a line in the output image with the maximum image value Parameter Information direction Type int Default 0 Values N 0 NW 1 W 2 SW 3 S 4 SE 5 E 6 NE 7 optimization Type int Default 1 Values preciseness 0 speed 1 Demo iGraph Operators Filters compass igr C Prototype CorOpRtn cor_compass CorImage In CorImage Out int direction int optimization 79 CorOpRtn cor_compass_consume CorImage In int direction int optimization Headers include WITHOME h wFilter h Libraries WITHOME lib wObj lib WITHOME lib wFilter lib convid General one dimensional image convolution Description The convld operator applies one dimensional convolution by a user specified kernel to each row of the input im
308. int edgeType int subpixel float lineLength int numLines int lineWidth int smoothing Headers include WITHOME h wEdges h Libraries WITHOME lib wObj lib WITHOME lib wEdges lib 74 getContours Extract contours from a skeletonized binary image Description The getContours operator extracts contours represented by chains of pixels from a binary image The operator s output is a vector of geometric polyline objects in which each polyline represents a chain of contiguous 8 connected nonzero valued pixels in the input image No heuristics are used to ensure qualities other than connectedness in the chains The polylines produced by getContours can be processed by the geom2lines operator which converts the polyline into straight line segments These lines can be further processed with the mergeLines operator Demo iGraph Operators Edges FindCircle C Prototype CorOpRtn cor_getContours CorImage In CorGeomVector Contours Headers include WITHOME h wEdges h Libraries WITHOME lib wObj lib WITHOME lib wEdges lib getXEdges Find edges crossing a specified path in the image ae ae Description The getXEdges operator finds step or peak edges along a given path in image The inputs to the operator are an integer or float image at the upper input port and graphic or geometric object which specifies the path at the Path input port The path may be of any geometric type except
309. ir constant and cat C Prototype CorOpRtn cor_concat CorObj Ini CorObj In2 CorObj Out Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib concat Concatenate objects tet Description The concat operator concatenates two input objects The objects must be either vectors integers files or strings and both objects must be the same type The nature of the concatenation depends on the type of the input objects If the input objects are vectors the elements of the second vector from the In2 input port are added to the end of the first vector from the In1 input port The elements of the two vectors must be the same type If integers are given then the output is a Point with the integer from the Ini input port the x coordinate and the integer from the In2 port the y coordinate If strings or files are presented as inputs then the output is a concatenated string For example you may want to build up a file name with using splitDir constant and concat C Prototype CorOpRtn cor_concat CorObj In1 CorObj In2 CorObj Out Headers include WITHOME h wObjman h 155 Libraries WITHOME lib wObj lib WITHOME lib wObjman lib const Generates a basic constant object when triggered aki Description The const operator generates a constant object every time an input is received If the type parameter is blank the value and type
310. is controlled by the sensitivity parameter The Range mode allows the user to specify the angular range through which the characters in the image may be rotated The range of angles is specified by the minAngle and maxAngle parameters The angle range is used to limit the rotation of the rotation sensitive features before matching rather than to reject matches in which the best fit is rotated outside the range This means that characters that are rotated outside the specified range by a small amount may still be considered matches This mode of operation is somewhat slower than the other two modes The featuresUsed parameter can be used in this mode to speed up matching sensitivity The sensitivity parameter controls the sensitivity of matching to rotation when rotation parameter is set to Non Invariant The parameter controls the relative weight of rotation sensitive and rotation invariant features used The parameter is set to a value from 0 to 1 where 0 specifies rotation invariant features only and 1 specifies rotation sensitive features only Lower values of this parameter tend to allow greater tolerance for rotation in the image however it is difficult to quantify and may vary among the different characters in the font Lower values also tend to provide more robust discrimination between characters with simple perimeter shapes particularly those exhibiting some symmetry Consequently this parameter is usually set to the
311. is set to 0 the polygons are positioned so that the number of rows and columns are approximately equal Parameter Information width Type int Default 0 height Type int Default 0 offset Type int Default 2 Values None 0 Tile 1 Original 2 cols Type int Default 0 C Prototype CorOpRtn cor_ocrPlotFont CorFontModel FontModel CorGraphicVector Perimeters int width int height int offset int cols 246 Headers include WITHOME h wOcr h Libraries WITHOME lib wObj lib WITHOME lib wOcr lib ocrRoiBin Binarization using ROI based thresholding to ual Description The ocrRoiBin operator binarizes regions of interest ROIs in a text image based on either supplied thresholds or thresholds calculated from the pixel values in the ROIs Inputs In The input image may be any integer image type or a float image ROIs The ROIs input can either a single rectangle Graphic or Geometry type object or a vector of rectangle Graphic or Geometry objects Outputs Out The operator produces a binary unsigned 8 bit image with white text on a black background within the ROIs and black background outside the ROIs Threshold The operator outputs a vector of float values that holds the thresholds used in the ROIs A vector of thresholds is often calculated in an initialization step or in a first pass for a set of ROIs The vector is then supplied as an input parameter for subsequent applications of
312. is specified by the offsetMethod parameter The low and high offset values are determined using the loOffset and hiOffset parameters respectively When the offsetMethod parameter is set to absolute the loOffset and hiOffset parameter values are simply added to the local mean value to determine the low and high threshold values In the other modes the parameters are used as multiplying factors to determine the threshold offset values In the mean min mode the difference between the local mean pixel value and the image mimimum pixel value is multiplied by the parameter values and the resulting offset values are added to the local mean In the max mean mode the difference between the image maximum and the local mean is multiplied by the parameter values to determine the offsets In the standard deviation mode the offsets are determined by multiplying the standard deviation of the pixel values in the neighbourhood by the parameter values Note that the standard deviation is calculated over the same neighbourhood as the local mean value while the minimum and maximum used in the previous two modes described are the minimum and maximum over all pixel values in the image The minOffsetBase parameter is used in the mean min max mean and standard deviation modes to help suppress noise problems in low signal regions of the image The parameter value specifies the minimum value of the difference in the first two modes or the standard devi
313. ition on frame grabber a Description The acqStart operator configures the current frame grabber for continuous acquisition and starts frame transfer to a circular queue of image buffers This is a generalization of so called ping pong buffering in which uses two buffers only For some hardware the queue size is fixed while for others the queueSize parameter can be used to set the number of image buffers in the queue If the queue size is user modifiable then the number of buffers in the queue is limited only by available memory After acqStart has executed the acquire operator immediately returns the image from the head of the queue unless the queue is empty in which case it waits for a frame as usual This allows processing to proceed in parallel with the transfer of image data to memory As long as the average processing time per frame does not exceed the interval between frames delivered by the frame grabber then no frames are dropped as long as the queue is large enough Clearly if there is great variability in the time taken to process individual frames then the queue needs to be larger If the frame grabber delivers frames quickly enough to fill the queue then subsequent frames will be dropped In this case the next acquire operator will return in the 301 dropped output the number of frames dropped since the previous execution of acquire Normally acqStart is called once and then a loop is entered in w
314. its parameter and the operator returns When the hardware counter value reaches the value given by the count parameter the selected output lines are switched to the values given by bits and remain so until the hardware counter value is equal to count duration At that time the selected output lines are switched back to the values they had before count was reached The acqloSetup operator may be executed before acqloWrite to configure the hardware appropriately If this is not done then the configuration will occur the first time acqloWrite is executed which may incur a performance penalty See the help for the acqIoSetup operator for more information on I O support Parameter Information port Type uint Default 0 mask Type uint Default 255 bits Type uint Default 255 296 mode Type int Default 0 Values immediate 0 delayed 1 count Type uint Default 0 duration Type uint Default 1 C Prototype CorOpRtn acqIowWrite uint port uint mask uint bits int mode uint count uint duration Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqlssueTrigger emulate a trigger Description Issue a trigger internally This is especially useful when using the frame grabber emulator to emulate externally triggered acquisition Demo iGraph Operators Video Acquisition trigger C Prototype CorOpRtn acqIssueTrigger Headers in
315. l igr C Prototype CorOpRtn cor_fitTwoLinesToPts CorObj Pts0 CorObj Pts1 CorGraphic Line0d CorGraphic Linel int lines Headers include WITHOME h wFit h Libraries WITHOME lib wObj lib WITHOME lib wFit lib gCentroid Compute weighted centroid of an ROI Description The gCentroid operator determines the centroid of a given region of interest ROI based on the values in the input image The centroid is calculated by a center of mass type calculation with the x coordinate xc given by xc sum fi xi sum fi where fi is the value and xi is the x coordinate of pixel i and the sums are over all pixels i in the ROI The y coordinate is calculated in the same way The input ROI can be either a graphic object or by a binary image Graphic object ROIs are rasterized and all pixels in the input image corresponding to the rasterized pixels including boundary pixels are 103 used in the calculation of the centroid When a binary image ROI is used all pixels in the input image that correspond to non zero pixels in the ROI image are used The ROI used for the centroid calculation is the intersection of the input ROI and the image ROI The graphic objects may be generated by the getData and createRect operators or may be constructed using the createGraphics and putField operators Binary images are often produced by one of the thresholding operators thresh iThresh or localThresh Demo iGraph
316. l image space using the matchSize operator with the sizeOp parameter set to center B on A Parameter Information direction Type int Default 0 Values forward 0 reverse 1 startAngle Type int Default 0 endAngle Type int Default 179 angleIncr Type float Default 1 accumulate Type int Default 0 Values max 0 sum 1 incr 2 Demo iGraph Operators Geometry convexHull igr C Prototype CorOpRtn cor_hough 107 CorImage In CorImage Out int direction int startAngle int endAngle float angleIncr int accumulate Headers include WITHOME h wFit h Libraries WITHOME lib wObj lib WITHOME lib wFit lib houghDir Hough transform with direction Description Hough transform with direction Parameter Information tolerance startAngle endAngle angleIncr accumulate degrees 108 Type int Default 5 Type int Default 128 Type int Default 127 Type float Default 1 Type int Default 0 Values max 0 sum 1 incr 2 Type int Default 0 Values 360 0 256 1 Demo iGraph Operators Geometry convexHull igr C Prototype CorOpRtn cor_houghDir CorImage Magnitude CorImage Orientation CorImage Out int tolerance int startAngle int endAngle float angleIncr int accumulate int degrees Headers include WITHOME h wFit h Libraries WITHOME lib wObj lib WITHOME lib wFit lib houghDirRoi Hough transform
317. l in the neighbourhood of p as defined by the kernel dimensions is on and the corresponding entry in the kernel is also on then the output pixel corresponding to p will be set In other words the kernel defines a mask which is layed over the input image with its center aligned on a pixel at row r and column c The output pixel at r and c will be set if any input pixel is set and the corresponding bit in the kernel mask is also set For example oorr row OrPrRrF Oct orroo oRomonome 1 Kernel 011 Output 00000 Ae Le 040 devel elt J 40 OF ed See 030 O 0 O80 0 2 Kernel tag Output 00000 Ore aL lO OnT NE xl 0 70 T F 0 0 7000 3 Kernel 1 0 Output Qt 28 0 TETE 0 Uy eae ao a 6 142 a Be awi e A an a A a fe Ee aad oe In the first example the output is only grown to the left of the input object because the kernel when aligned with any input pixel will set the output only if either the pixel corresponding to kernel center is set or the pixel one to the right of that input pixel is set In a similar fashion the second example grows the object to the right The third example uses a kernel of size 2x2 with its center being the top left bit in the kernel The resulting output object has been grown to the left and top of the input object The implementation of dilation used in WiT is defined in the paper by Haralick Sternberg and Zhuang IEEE Transactions on Pattern Analysis and
318. l lib extractRoi Extract an image s ROI Description The extractRoi operator extracts a sub image of the input image corresponding to the input image s intrinsic ROI The image ROI can be set using the setRoi operator C Prototype CorOpRtn cor_extractRoi CorImage In CorImage Out Headers include WITHOME h wPixel h Libraries WITHOME lib wObj lib WITHOME lib wPixel lib fill Fill a region of an image with a value 5 tort Description The fill operator fills a region in the input image with a pixel value specified by the fillValue parameter The region to be filled is specified by the mask parameter The mask parameter may be a Graphic or Geometric object or an 8 or 16 bit unsigned image a vector of Graphic or Geometric objects or an 8 or 16 bit unsigned images or a Blob vector The fill parameter is used to specify if the Foreground areas pixels inside the mask or the Background areas pixels outside the mask are filled Parameter Information mask Type CorObj fillValue Type CorObj Default OBJ_B T int 0 fill Type int Default 0 Values Foreground 0 Background 1 C Prototype CorOpRtn cor_fill_1 CorImage In CorImage Out CorObj mask CorObj fillValue int fill CorOpRtn cor_fill_1_consume CorImage In CorObj mask CorObj fillValue int fill Headers include WITHOME h wPixel h Libraries WITHOME lib wObj lib WITHOME lib wPixel
319. lay the data Placement sub panel Placement If Fix position is checked values of the X and Y properties are used to position the data window Otherwise the current automatic placement rules are used X The X position of the data window if Placement is set to Fix position Y The Y position of the data window if Placement is set to Fix position Property Panel Show Enable disable histogram display on histogram ramp Min Hist max Control scale of histogram displayed on ramp The larger the scale the taller the histogram will appear Default value is maximum value of histogram Ramp Similar to the ramp on the image property panel af Accept the current thresheld image and send it to the output z ort the iEqualize session No output FN Abort the iEquali ion N p is produced Parameter Information name Type CorString Default iEqualize Placement Type int Default 0 Values bitwise OR Fix position 1 X Type int Default 0 Y Type int Default 0 iKey Prompt the user for a keyboard character 59 Description The iKey operator is used when the user needs to be prompted for a single character from the keyboard during igraph execution The input object triggers this operator to create a pop up window which requires the user to hit a key on the keyboard As soon as the key is hit an output object a string with a single character is produced with the key as the first
320. lectAverage C Prototype CorOpRtn cor_addNoise CorImage In CorImage Out CorUshortVector NextSeed int affectedArea int noiseType float noiseAmount int useSeed CorUshortVector seed int reset Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib aluOp Perform arithmetic and logic operations on two images Description The aluOp performs binary arithmetic and logical operations given two input images A and B The input images may be grayscale or color images If both input images are grayscale images they may be any grayscale type and the output image type may be selected using the outputType parameter The default output image type is selected according to the following criteria 11 Input Image Types Output Image coma if either lif either input is float is float output is float is float S if either input is 16 output is 16 bit lif either input is signed either input is lif either input is signed output is output is signed Color input images may be combined with any type of grayscale image or with a color image of the same type RGB with RGB HSV with HSV and Yuv with Yuv If both input images are color images the operations are applied channel by channel If one of the images is a grayscale image its pixel values are applied to each channel of the color image If either input image is a color image the default output type
321. lected the value is that of the nearest pixel that is the pixel within which the sample point falls If the interpolate option is selected the value is determined by bilinear interpolation from the surrounding pixel values Parameter Information operation Type int Default 0 Values rasterize 0 sample 1 sampling Type int Default 0 Values interpolate 0 nearest 1 sampleSpacing Type float Default 1 C Prototype CorOpRtn cor_getPath CorImage In CorObj Path CorFpointVector Points CorFloatVector Values int operation int sampling float sampleSpacing Headers include WITHOME h wPixel h Libraries WITHOME lib wObj lib WITHOME lib wPixel lib getPixel Get a pixel value from an image Description The getPixel operator retrieves a single pixel of data corresponding to a pixel value in the input image The pixel value is located by the x y coordinate given in the xy parameter Parameter Information xy Type CorPoint Default 0 0 C Prototype CorOpRtn cor_getPixel CorImage In CorObj Pixel CorPoint xy Headers include WITHOME h wPixel h Libraries WITHOME lib wObj lib WITHOME lib wPixel lib getRow Get a row of pixels from an image Description The getRow operator retrieves the vector of pixel data corresponding to the row of pixels in the input image The row is given in the row parameter Parameter Information row Type
322. light background Roi A single ROI may be specified by a rectangular Graphic or Geometry object Multiple ROIs can be specified by a vector of Graphic or Geometry objects Outputs Decoded If the ROI is a single object then the output is a String object If a valid barcode is found the output string contains the characters represented by the barcode including the check character If no valid barcode is found an empty string is output If the ROI is a vector of rectangles then the operator seeks for a barcode in each ROI The output is a vector of String objects If a valid barcode is found in a ROI the corresponding element of the output vector contains the decoded string If no valid barcode is found in a ROI the corresponding output vector element is the empty string Parameters noise The noise parameter specifies the minimum low to high or high to low change in pixel value necessary to constitute the edge of bar minLength The minLength parameter specifies the minimum amount of characters in the decoded string Parameter Information noise Type int Default 5 minLength Type int Default 0 Demo iGraph Operators Matrix Codabar igr C Prototype CorOpRtn cor_getCodabar CorImage Target CorObj ROI CorObj Decoded int noise int minLength Headers include WITHOME h wMatrix h Libraries WITHOME lib wObj lib WITHOME lib wMatrix lib getCode128 Finds and decodes Code128 bar cod
323. line number offset outputs of operators such as webWaterfall and webLineCounter may be used to provide the input value The Color subpanel The color of the output graphics can be controlled using parameters on the Color 283 subpanel If the colors parameter is set to custom the remaining parameters can be used to specify whether the output graphics are outlines or filled graphic objects and to specify the color of the objects The default graphics are yellow outlines The label field of each graphic object is set to the id number of the associated element of the feature vector Parameter Information featureType Type int Default 0 Values centroids 0 major axis 1 bbox 2 perimeter 3 yOffset Type int Default 0 colors Type int Default 0 Values default 0 custom 1 outline Type int Default 0 Values off 0 on 1 penR Type int Default 0 Values 0 255 penG Type int Default 0 Values 0 255 penB Type int Default 0 Values 0 255 fill Type int Default 0 Values off 0 on 1 fillR Type int Default 0 Values 0 255 fillG Type int Default 0 Values 0 255 fillB Type int Default 0 Values 0 255 C Prototype CorOpRtn cor_webPlotFeatures CorObj Features 284 CorObj Graphics int featureType int yOffset int colors int outline int penR int penG int penB int fill int fil1R int fillG int fi11B Headers include WITHOM
324. ll icon which E makes it more convenient to use in large igraphs See Also display Parameter Information name Placement X Y Maxw MaxH MinW MinH Range RangeLow RangeHigh Equalize 66 Type CorString Default display Type int Default 0 Values bitwise OR Fix position 1 Max size 2 Min size 4 Type int Default 0 Type int Default 0 Type int Default 0 Type int Default 0 Type int Default 0 Type int Default 0 Type int Default 0 Values Auto 0 User 1 Type float Default 0 Type float Default 255 Type int Default 2 EqLow EqHigh XScale YScale Inspect Track Colormap ColormapName Format FontFamily Style Size TextRed TextGreen TextBlue BackRed Values Actual 0 User 1 Max 2 Type float Default 0 Type float Default 255 Type float Default 1 Type float Default 1 Type int Default 0 Values bitwise OR X Profile 1 Y Profile 2 Magnifier 4 Type int Default 0 Values Mouse drags 0 Mouse moves Type int Default 0 Values Grayscale 0 Pseudo color 1 Shared 2 Custom 3 Type CorFile Type int Default 0 Values Yes 0 No 1 Type CorString Default Type int Default 0 Values Normal 0 Italics 1 Bold 2 Bold Italics 3 Type int Default 11 Type int Default 0 Values 0 255 Type int Default 0 Values
325. lowest value that provides reliable discrimination between rotated or reflected character pairs minAngle and maxAngle These parameters specify the angular range in the Range rotation mode The parameter values are given in degrees and are not limited in range negative values and values greater than 360 are valid The angles are evaluated modulo 360 so a minAngle value of 0 and a maxAngle value of 370 will give a 10 degree range not a 370 degree range The exception to this is when the values indicate the same angle If the actual parameter values are the same the angle range is 0 if the actual parameter values are different by a multiple of 360 the range is 360 degrees featuresUsed The featuresUsed parameter controls the fraction of the feature set used in the Range mode The parameter must be set to a value from 0 to 1 Settings of or near 0 will result in the use of minimal feature set All the features will be used when the parameter is set to 1 Lower values increase the operator s speed but also reduce the ability to discriminate between different characters scaleInvariant When this parameter is set to Yes the operator automatically normalizes the characters for scale It should be noted that very small characters can suffer from pixellization effects that can distort the shape of the character perimeter 243 244 When set to No the operator limits the range of scale corrections that can be applied to the character
326. ls and type in vectors The data field refers to the pointer to memory containing the elements and is most often used with an index value to specify one element in the vector The field name data n specifies element n of the vector An error occurs if there is no element n in the vector Field names of extended or user defined types can be seen by viewing an object of the type using a probe or the display operator For example displaying a Feature object produces a list of the feature s fields id xmin xmax and their values The names as displayed including case can be used to access these fields Substructures embedded within structures can be specified using either or gt notation Fields can be nested to an arbitrary depth For example the Graphic object has a field called geom which is itself an object that includes the field plist This field can be accessed by specifying geom plist or geom gt plist as the field parameter The plist field is the vector of points that describe the graphic object The number of elements in the plist vector can be replaced by specifying geom plist size The coordinate values of a given point in the graphic object can be accessed using the vector notation described above For example geom plist data 0 x specifies the x coordinate of the first point in the vector The putField operator cannot allocate memory for an empty field The most common instance in which this arises is when
327. lt 255 0 0 Placement Type int Default 0 Values bitwise OR Fix position 1 X Type int Default 0 Y Type int Default 0 width Type int Default 500 height Type int 70 Default 500 Range Type int Default 0 Values Auto 0 User 1 RangeLow Type float Default 0 RangeHigh Type float Default 255 Equalize Type int Default 0 Values Auto 0 User 1 EqLow Type float Default 0 EqHigh Type float Default 255 Colormap Type int Default 0 Values Grayscale 0 Pseudo color 1 Custom 2 ColormapName Type CorFile volume Display volume rendering of a vector of images Description The volume operator computes and displays a volume rendering of a vector of images in which each image represents a slice of a 3D volume The color and opacity of rendered objects are determined from the pixel values and can be changed interactively by the user The user can also interactively control the viewpoint and a cut plane Rendered images can be saved in the WiT image format The input must be a vector of 8 bit unsigned images in which each image represents one slice of a volume All images in the vector must be the same size The volume operator assumes the volume sampling is equal in all three dimensions If this is not the case the vector of images must be preprocessed to equalize the sampling or the resulting rendered volume will be distorted The output of the operator is a color image as
328. lt 0 targetY Offset Type int Default 0 searchRange Type int Default 5 result Type int Default 2 Values center 0 upper left 1 bounding box 2 vicinity Type int Default 5 red Type int Default 255 Values 0 255 green Type int Default 0 Values 0 255 blue Type int Default 0 Values 0 255 Demo iGraph Operators Find TMatch 97 C Prototype CorOpRtn cor_tmatch CorImage target CorObj Template CorObj ROI CorIntVector TemplateNumber CorGraphicVector Positions CorFloatVector CorrValue int method int correlation float threshold float sufficient_threshold int numToMatch int templateSampling int templatexOffset int templateYyOffset int targetSampling int targetxXOffset int targetYOffset int searchRange int result int vicinity int red int green int blue Headers include WITHOME h wFind h Libraries WITHOME lib wObj lib WITHOME lib wFind lib tmatchSparse Template matching Description The tmatchSparse operator performs template matching using a template specified as a vector of points and a vector of corresponding pixel values The TemplateCoor input must be a vector of Point objects indicating the point locations in the template that will be evaluated The templateVal input is a vector of numeric values that represent the template values at the points specified in the 98 templateCoor vector This allows the specification
329. mage by selecting Max The parameters on the Rows subpanel control duplication and spacing of rows in the output image If the dataThickness parameter is set to a value greater than one each element vector will be copied into that number of adjacent rows in the output image The gapThickness parameter allows the user to specify the number of rows of zero valued elements that separate the row or rows copied from each element vector Again this may result in inconsistent data values for compound data types If the addLastGap parameter is set to No the gaps will only be added between the row or groups of rows copied from the element vectors If it is set to Yes the gap will also be added after the last copied row Parameter Information setRoiWidth Type int Default 1 Values Min 0 Max 1 dataThickness Type int Default 1 gapThickness Type int Default 0 addLastGap Type int 52 Default 0 Values No 0 Yes 1 Demo iGraph Operators Blobs GetBlobFeatures igr Operators Blobs GetBlobFeaturesMasked igr Operators Blobs SelectFeatures igr C Prototype CorOpRtn cor_vectorsToImage CorVectorVector Vectors CorImage Image int setRoiWidth int dataThickness int gapThickness int addLastGap Headers include WITHOME h wConvert h Libraries WITHOME lib wObj lib WITHOME lib wConvert lib Display Display and interactive operators clearDisplay Clear all pop up displays Description
330. mage is then filtered using the gaussian filter specified by the filter parameter The filters are described in detail in the gauss operator within the pyramid library If the None option for filter type is indicated then no filtering is performed The outputWidth and outputHeight parameters allow the user to specify the width and height of the output image However the values input are verified to assure that the resulting image will have an appropriate size If the reduce operator is applied to an image I 1 and used to create an image I i 1 then the output width and output height should be set to the dimensions of image I i Parameter Information filter Type int Default 0 Values None 0 2x2 1 3x3 2 5x5 3 7x7 4 outputWidth Type int Default 0 outputHeight Type int Default 0 C Prototype CorOpRtn cor_expand CorImage In CorImage Out int filter int outputWidth int outputHeight Headers include WITHOME h wPyramid h Libraries WITHOME lib wObj lib WITHOME lib wPyramid lib 186 fsd Generate a pyramid using FSD method 2 n a Pis Description The fsd operator generates two image sequences G i and L i The sequence G i that is output on the top port is a Gaussian pyramid and is identical to that which would be generated by the gaussPyramid operator The sequence L i is a Laplacian pyramid and is output on the bottom port Each image L i is a difference between image
331. mages Parameter Information structElem Type Corlmage Default 0 0 0 0 0 0 0 0 0 Demo iGraph Operators Morphology skeleton igr C Prototype CorOpRtn cor_thicken CorImage In CorImage Out CorIntImage structElem CorOpRtn cor_thicken_consume CorImage In CorIntImage structElem Headers include WITHOME h wMorpho h Libraries WITHOME lib wObj lib WITHOME lib wMorpho lib thin Binary thinning filter as rE t r Description The thin operator performs a thinning type binary pattern matching operation on an image If the pixels in the neighbourhood of a given pixel match the pattern of pixels specified in the kernel then the corresponding pixel in the output image is set to zero otherwise it is unchanged The kernel parameter allows the user to specify the pattern Kernel elements set to 0 match 0 value pixels in the input image kernel elements set to 1 or any positive integer match non zero pixels in the input image and kernel elements set to 1 or any negative integer match all pixel values in the input image The input and output of the thin operator are both 8 bit unsigned images This operator could be executed in two modes The user may choose the mode by using the optimization parameter If the parameter value is speed then execution goes way faster otherwise operator delivers more precise results Parameter Information structElem Type Corlmage Default
332. mber of regions The pixels not assigned to any region are given a value of zero It should be noted that these zero pixels do not necessarily form a continous dividing line between the regions The connected parameter controls the neighbourhood searched for descending paths from a pixel It also affects the interpretation of distance when dividing plateau areas among regions Setting this parameter to 4 will generate four connected basins and the distances on plateaus are city block distances the distance measured moving in vertical and horizontal directions only Selecting 8 will generate eight connected basins and use a chess board distance where all eight connected pixels are considered to be one unit length away Parameter Information connected Type int Default 1 Values 4 0 8 1 Demo iGraph Operators Segmentation Watershed C Prototype CorOpRtn cor_watershed CorImage In CorImage Out int connected Headers include WITHOME h wSegment h Libraries WITHOME lib wObj lib WITHOME lib wSegment lib zeroX Zero crossing detector E a Sa Description The zeroX operator finds zero crossings in the rows of a signed 8 bit integer signed 16 bit integer or float input image One of three types of zero crossings may be found all lo gt hi or hi gt lo The operator produces a binary unsigned 8 bit image marking the locations of the zero crossings found in the input image A low to
333. mo iGraph Operators Calibration calibrate igr C Prototype CorOpRtn cor_restorelImage CorImage image CorCalibrateInfo calibrateInfo CorImage transformedData int mode Headers include WITHOME h wCalibration h Libraries WITHOME lib wObj lib WITHOME lib wCalibration lib Color The Color library provides support for manipulating color images The library includes operators that convert between color formats and operators that convert between single and multi channel images bayer Perform conversion between RGB and Bayer format E Description The bayer operator converts images between the Bayer color image format and RGB The Bayer format assigns each pixel in an unsigned eight bit image the value of one color channel RGB images are created by interpolating neighbouring pixel values to get the two missing color channels at each 34 pixel RGB images are sampled to create Bayer format images Pixels in a row of a Bayer image alternate between the green channel value and either the red or the blue channel value as shown below RGB images are converted to the Bayer format by sampling the appropriate color channel value at each pixel When converting from the Bayer format to an RGB image the missing color channel values are determined by linear interpolation using neighbouring pixel values for the color channel in question If the method parameter is set to Linear the interpolated valu
334. mo iGraph Operators Geometry convexHull igr C Prototype CorOpRtn cor_getBoundingBox CorObj In CorGraphic Out Headers include WITHOME h wFit h Libraries WITHOME lib wObj lib WITHOME lib wFit lib 106 getVertexPts Get vertex points of a geometric object or point set Description The getVertexPts operator finds the vertices of a single Graphic Geometry Edge or Line object or a vector of Graphic Geometry Edge or Line objects The output is a vector of Real Points For point line polyline polygon and text type geometric objects the operator produces the object s point list In the case of text objects only the text reference position the lower left corner of the text string is used For rectangles the operator produces the points corresponding to the upper left upper right lower right and lower left corners For circles the operator outputs the four points corresponding to the minimum and maximum x and y points on the circle For arcs the operator produces the points corresponding to the minimum and maximum x and y points on the arc and the endpoints of the arc Demo iGraph Operators Geometry convexHull igr C Prototype CorOpRtn cor_getVertexPts CorObj In CorFpointVector Out Headers include WITHOME h wFit h Libraries WITHOME lib wObj lib WITHOME lib wFit lib hough Hough transform Description The hough operator computes the forward or reverse
335. mposite Component is found in a ROI the corresponding element of the output vector contains the decoded string If no valid Composite Component is found in a ROI the corresponding output vector element is the empty string Parameters noise The noise parameter specifies the minimum low to high or high to low change in pixel value necessary to constitute the edge of bar type The type parameter specifies the Composite Component barcode type If parameter value is regular then Composite Component belongs to the RSS 14 RSS 14 Truncated RSS 14 Stacked or RSS 14 Stacked Omnidirectional otherwise it belongs to RSS 14 Limited Parameter Information noise Type int 222 Default 5 type Type int Default 0 Values regular 0 limited 1 Demo iGraph Operators Matrix BC412 igr C Prototype CorOpRtn cor_getCCA CorImage Target CorObj ROI CorObj Decoded int noise int type Headers include WITHOME h wMatrix h Libraries WITHOME lib wObj lib WITHOME lib wMatrix lib getCodabar Finds and decodes Codabar bar code j a HIM Description The getCodabar operator accepts a binary or grayscale image and one or more rectangular regions of interest ROIs then searches the ROIs for Codabar barcodes If a barcode is found in a region it is decoded to a string Inputs Target The input image must be an unsigned 8 bit grayscale image with the barcode represented as dark bars on a
336. mum size in at least one dimension The filter parameter specifies the type of gaussian filter used during image reduction The filters correspond to those provided by the gauss operator within the filter library The minsize parameter denotes the minimum size of image that will be accepted as input to the operator Parameter Information filter Type int Default 0 Values 2x2 0 3x3 1 5x5 2 7x7 3 minsize Type int Default 8 C Prototype CorOpRtn cor_fsdStep CorImage In CorImage Gaussian CorImage Laplacian int Final int filter int minsize Headers include WITHOME h wPyramid h Libraries WITHOME lib wObj lib WITHOME lib wPyramid lib gaussPyramid Generate a gaussian pyramid 7 Description The gaussPyramid operator generates a sequence of images corresponding to a gaussian pyramid Each image G i is a reduction of image G i 1 Notationally input image reduce G i 1 for i gt 1 QAQ The filter parameter selects the type of gaussian filter to be applied The effect of each filter is described in the WiT filter library during the discussion of the gauss operator The minsize parameter defines the minimum size that an image G i can be before a reduce operation will not be performed and pyramid generation will terminate Parameter Information filter Type int Default 0 Values None 0 2x2 1 3x3 2 5x5 3 7x7 4 minsize Type int Default 8 C Prototype CorO
337. n There are five case insensitive variables available for access by the user Aora top input Borb bottominptt i Xorx current vector column value Yory current vector row value Kork the value specified by userConstant The available operators ordered by precedence from highest to lowest are listed below and may be used to form the expression SS exponentiation multiplication division aia addition subtraction subtraction lt lt gt gt bit shift left bit shift right lt gt boolean comparisons l e 1 bitwise and or xor and or xor amp amp boolean conjunction boolean disjunction Unary operators a unary negation boolean not abs sqrt absolute value square root exp sin cos trigonometric functions tan asin acos inverse trigonometric atan functions log 10g10 natural log log base 10 log nega base 2 floor ceil round floor ceiling and rounding functions rint may be used in place of round General Conditional operator comparison result if true result if false Precedence parentheses can be used to alter precedence Note that all calculations are performed using double precision except for the modulus and bit operations which have their operands cast to integer first To perform integer division rather than floating point division use floor A B
338. n the edges of the regions in the input image 145 Parameter Information kernel Type Corlmage Default 1 1 1 1 1 1 1 1 1 Demo iGraph Operators Morphology hitormiss igr C Prototype CorOpRtn cor_ldilate CorImage In CorImage Out CorIntImage kernel CorOpRtn cor_ldilate_consume CorImage In CorIntImage kernel Headers include WITHOME h wMorpho h Libraries WITHOME lib wObj lib WITHOME lib wMorpho lib lerode Labeled image erosion Description The lerode operator erodes labeled images maintaining image pixel values in the eroded images The effect of this operator is the equivalent to applying the berode binary erosion operator to the pixels at each distinct non zero pixel value in the image and assigning the pixel value to the active pixels in the output image The result is that the equal valued regions in the image are eroded with adjacent regions with different pixel values being separated 146 Parameter Information kernel Type Corlmage Default 1 1 1 1 1 1 1 1 1 Demo iGraph Operators Morphology hitormiss igr C Prototype CorOpRtn cor_lerode CorImage In CorImage Out CorIntImage kernel CorOpRtn cor_lerode_consume CorImage In CorIntImage kernel Headers include WITHOME h wMorpho h Libraries WITHOME lib wObj lib WITHOME lib wMorpho lib median Median filter Description The median operator performs medi
339. nce between A and B min minimum of A and B max maximum of A and B Each pixel of the output is determined by performing the indicated operation on the corresponding pixels in image A and image B The expression field allows the user to specify a C style mathematical equation which will determine the output image The top input image can be referenced in this equation using the variable A while the bottom image is given using B The result for each pixel can be affected by that pixel s particular column and row using the X and Y variables respectively For example the expression Y gt 5 A B will produce an output image where the first 5 rows are identical to the first five rows of B Otherwise the output image corresponds to A For more on the expression syntax and use see calc Parameter Information aluOp Type int Default 0 Values A B 0 A B 1 A B 2 A B 3 OR 4 AND 5 XOR 6 src 7 overlay 8 expr 9 abs A B 10 min 11 max 12 sizeOp Type int Default 0 Values clip B 0 expand A 1 overlayMask Type CorHex Default Oxffff offset Type CorPoint Default 0 0 expression Type CorString Default A outputType Type int Default 0 Values default 0 8 bit unsigned 1 8 bit signed 2 16 bit unsigned 3 16 bit signed 4 float 5 Demo iGraph Operators Calculate AddNoise Operators Calculate CollectAverage C Prototype CorOpRtn cor_aluOp CorImage A CorImage B CorImage Ou
340. nd 172 field value such as lines of text characters under small rotations Parameter Information fieldl Type CorString Default y direction Type int Default 0 Values ascending 0 descending 1 tolerance Type float field2 Type CorString Default x direction2 Type int Default 0 Values ascending 0 descending 1 output Type int Default 0 Values object 0 index 1 C Prototype CorOpRtn cor_sortTwoFields CorVector InVector CorVector SortedVector CorString fieldl int directionl float tolerance CorString field2 int direction2 int output Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib swapBytes Swap byte order in two byte images a E Bg ite Description The swapBytes operator takes an image as input and outputs an image of the same type but has the low and high bytes words swapped The operator is useful when images are being processed on different architectures On PCs an ordinary C int variable is stored in memory starting from the least significant byte up to the most significant byte On the other hand Sun Workstations store the same variable in memory starting from the most significant byte down to the least significant byte For RGB and 8 bit images swapBytes just outputs the same image as the input image No swapping is necessary For 16 bit images the operator swaps the low and high by
341. ndicular lines to the two edge lines The parameters on the SearchBoxes subpanel are used to set the box around the end points of the reference line that will be searched for edges The search box is a rectangle oriented so that its height is parallel to the reference line and its width is perpendicular to the reference line A number of lines evenly spaced within the box and parallel to the reference line are searched for edges The boxWidth0 boxHeight0 and numPaths0 parameters set the search lines near the first end of the reference line and b gt boxWidth1 boxHeight1 and numPaths1 parameters set the search lines near the second end The parameters on the Edges subpanel control the edge search along the search lines The edgeTypeO parameter can be used to direct the search to look for rising falling or unknown edges The edge type is relative to the direction from the end point to the middle of the reference line If the type is set to unknown the operator will chose the direction which produces the strongest edges A minimum threshold can be set for edge slope using slopeThreshO and level change using levelThresh0 The edgeTypel slopeThresh1 and levelThresh1 parameters apply to the second end of the reference line The search returns the strongest edge point on each search line that meets both threshold criteria If no edge points are found at either end of the reference line an error condition occurs The subpixel parameter is u
342. ne is more than a specified error distance from the new polyline the most distant vertex from the new polyline is found and added as a vertex of the new polyline The process is repeated for the remaining vertices relative to the new polyline The error distance is specified in pixels using the max Deflection parameter The error distance can be either the maximum distance from any vertex to the new polyline or the mean distance from the new polyline to the original polyline The type of distance measure used is specified by the deflection Type parameter The mean distance is measured on only one side of the new polyline at a time The adjust parameter specifies which side of new polyline will be searched for vertices to add Specifying left or right will allow each successive search to be limited to one side of the line If the adjust parameter is set to both the operator will first consider the side with the greatest distance from the original estimate The algorithm will be applied until the error distance on that side of the approximation is within the specified limit It will then consider vertices on the other side of the new polyline When the new polyline is within the error distance on that side the first side may be considered again This process continues until either the distance on both sides is below the error distance or it has switched sides the number of times specified in the nlterations parameter Setting the nIterations pa
343. ne or more pixels in the image A small Pix sq gives you a fine and more accurate surface but takes longer to render So when you are adjusting your view angle distance etc it is a good idea to first set Pix sq to a large value do the viewing adjustments and then set Pix sq to the value you want Pix sq behaves slightly differently on a color and a monochrome display On a color display no boundaries are Wireframe drawn between the squares Boundaries are drawn with the size specified by Wireframe see below If Wireframe is less than Pix sq no wireframe is drawn On a monochrome display boundaries are drawn between the Pix sq squares and the Wireframe is not drawn except when Pix sq is set to zero When Pix sq is zero then for both color and monochrome displays the 3 D surface will no longer appear solid Instead a wireframe is drawn at the size of Wireframe The figure below shows such a view Specifies the size of the visible grid overlaid on the 3 D surface Wireframe behaves differently in color and monochrome displays See Pix sq above for details Parameter Information name background hAngle vAngle zoom Type CorString Default 3d surface Type int Default 0 Values No 0 Yes 1 Type float Default 45 Type float Default 45 Type int Default 65 scale grid wireframe wireColor Placement X Y width height Range RangeLow
344. ned on the top and bottom of the icon and the Gate control is fed from the left side Demo iGraph Operators clearObjs Geometry The Geometry library supports spatial transformations of images and other objects The library include operators that perform spatial transformations such as rotation shearing scaling and perspective warping on images and other objects and operators that perform spatial calibration functions flip Perform a simple geometric transforms on an image 123 Description The flip operator performs one of several common geometric transformations on its input image These include flipping in x and y transpose and rotations by multiples of 90 degrees The transformation type is specified by the op parameter flip y x X y Y y flip x x X x y y x pose x y y x r 90 x y y X x r 180 x X x y Y y r 270 uF Ysy y eS where x y is a pixel location in the output image x y is the location of the corresponding pixel in the input image and X and Y are the maximum x and y pixel coordinates in the input image Parameter Information op Type int Default 0 Values flip y 0 flip x 1 transpose 2 rotate 90 3 rotate 180 4 rotate 270 5 C Prototype CorOpRtn cor_flip CorImage In CorImage Out int op Headers include WITHOME h wGeom h Libraries WITHOME lib wObj lib WITHOME lib wGeom lib getRigidBodyTransform Produce r
345. nformation width Type int Default 3 Demo iGraph Operators Filter Entropy Operators Filter V ariance C Prototype CorOpRtn cor_lopass1d CorImage In CorImage Out int width CorOpRtn cor_lopass1d_consume CorImage In int width Headers include WITHOME h wFilter h Libraries WITHOME lib wObj lib WITHOME lib wFilter lib lopass2d Two dimensional smoothing Ax i i i Description The lopass2d operator applies two dimensional low pass filtering smoothing to the input image The effect of the smoothing operation is to blur the image reducing the sharpness of edges and fine detail in the image and reducing the effects of high frequency noise Both of these effects may be useful before segmentation or feature recognition is performed The smoothing operation is performed with a moving average kernel of a size in pixels specified by the width and height parameters The effect of lopass2d is equivalent to that of the conv2d operator with kernel size height by width all kernel values equal to 1 and a scale factor equal to the number of elements in the kernel array but the simple kernel structure allows a more efficient implementation This operator could be executed in two modes The user may choose the mode by using the optimization parameter If the parameter value is speed then execution goes way faster otherwise operator delivers more precise results The operator accepts integ
346. ng artifacts associated with the perfect low pass filter This operator could be executed in two modes The user may choose the mode by using the optimization parameter If the parameter value is speed then execution goes way faster otherwise operator delivers more precise results 83 Four different discrete approximations to gaussian filters are provided A filter of size 2x2 3x3 5x5 or 7x7 may be selected using the filterSize parameter Use of the gauss operator is equivalent to using conv2d with the following filters and scale parameters Type 2x2 Scale 4 Type 3x3 Scale 16 2 4 2 Type 5x5 Scale 256 4 6 4 4 16 24 16 6 24 36 24 4 16 24 16 4 6 4 PRO Wer Type 7x7 Scale 2048 0 So Lae Foni 5 0 5 25 S264 52 25 5 13 92702 126 102 82 413 16 64 126 156 126 64 16 13 52 LOZ 126 102 527 138 9 Zo 82 64 52 25 5 0 5 13 16 13 5 0 The operator accepts integer and float type input images and color images in the case optimisation parameter value is speed The output image is the same type as the input image Parameter Information filterSize Type int Default 0 Values 2x2 0 3x3 1 5x5 2 7x7 3 optimization Type int Default 1 Values preciseness 0 speed 1 Demo iGraph Operators Filter Entropy Operators Filter V arianceTexture C Prototype CorOpRtn cor_gauss CorImage In 84 CorImage Out int filterSize int optimization CorOpRtn cor_gauss_consume Co
347. nsforms CrossCorrelation C Prototype CorOpRtn cor_complexSplit CorImage Complex CorImage Real_Mag CorImage Imag_Phase int output Headers include WITHOME h wComplex h Libraries WITHOME lib wObj lib WITHOME lib wComplex lib spectrum Generate power spectrum of an image gs FM oe Description The spectrum operator converts a complex or float frequency domain image to a power spectrum representation The input image is generally produced by a forward fft operation or in the case of a float input image by applying the complexSplit operator the output of the fft The operator produces an 8 bit unsigned image that represents a logarithmic scaling of the input image The output image is an 8 bit unsigned image formed by taking the base 10 logarithm of the squared magnitude of the input image scaled to the range 0 to 255 Due to this scaling pixel values in the output image cannot be directly related to the input image values A value of 1 is added to the squared magnitude of each image pixel before the logarithm is taken to ensure that the result is non negative This will distort the spectrum near small magnitude values 39 The output image is the same size and has the same ROI as the input image Image values outside the ROI are undefined Demo iGraph Operators Transforms Transforms C Prototype CorOpRtn cor_spectrum CorImage In CorImage Out Headers include WITHOME h w
348. nt 5 float 6 double 7 conversion Type int Default 0 Values round 0 floor 1 ceiling 2 Demo iGraph Operators Calculate Lut C Prototype CorOpRtn cor_makeSeq CorIntVector Sequence CorString expression int size double userConstant int type int conversion Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib movingAverage Take the running average of a sequence of images F po Description The movingAverage operator produces an output image that is an average of the current input image and previous input images The way in which the operator calculates the average is controlled by the type parameter When the type parameter is set to Window the operator produces the mean of the last N images received by the operator where N is the smaller of the number specified using the number parameter and total number of images received When the type parameter is set to Weighted the output image is the sum of 1 N times the input image and N 1 N times the previous output image Again N is the smaller of the number specified 25 using the number parameter and total number of images received All input images to the operator must be the same image type and size and the ROIs must be the same The output image has the same type size and ROI as the input images Parameter Information number Type int Default 4 type Type int Default 0
349. nt image The clear parameter blanks the image rows stored from previous images in the stream after processing the current image This parameter is also used as an input parameter when it is required When a one is received the next image in the stream will have blank zero valued pixels in offset regions Parameter Information validRange Type CorVector offset reset clear Type CorVector Type int Default 0 Values No 0 Yes 1 Type int Default 0 Values No 0 Yes 1 C Prototype CorOpRtn cor_webRealign CorImage In CorImage Out int Offset CorPointVector validRange CorIntVector offset int reset int Headers clear include WITHOME h wWeb h Libraries WITHOME lib wObj lib WITHOME lib wWeb lib 285 webRowExtend Extend images in a stream for neighbourhood operations BAA ee a Description The webRowExtend extends an image from a stream of images by prepending rows from preceding images and appending a specified number of rows from subsequent images The operator is applied to images in a stream that represent vertically adjacent blocks from a single image space so that neighbourhood image processing methods such as filtering will produce correct results at the top and bottom of the images Images extended in this manner can be returned to their original size using the webRowRemove operator Inputs In The input images in a stream must
350. nt of the vector or list using the method parameter If the byElement option is selected then the element parameter is used to specify the position at which the element is added with the first element indexed by zero Selecting byElement with element set to zero is equivalent to selecting First Parameter Information method Type int Default 0 Values first 0 last 1 byElement 2 element Type int Default 0 C Prototype CorOpRtn cor_addElem CorObj Vector CorObj NewElement CorObj NewVector int method int element Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib cat Concatenate objects Description The cat operator concatenates two input objects The objects must be either vectors integers files or strings and both objects must be the same type The nature of the concatenation depends on the type of the input objects If the input objects are vectors the elements of the second vector from the In2 input port are added to the end of the first vector from the In1 input port The elements of the two vectors must be the same type If integers are given then the output is a Point with the integer from the In1 input port the x coordinate and the integer from the In2 port the y coordinate If strings or files are presented as inputs then the output is a concatenated string For example you may want to build up a file name with using splitD
351. nt optimization Headers include WITHOME h wMorpho h Libraries WITHOME lib wObj lib WITHOME lib wMorpho lib rank Rank filter Description The rank operator performs rank filtering on greyscale images The operator sets each pixel in the output image to the pixel value in the neighbourhood nearest to a specified percentile calculated over the neighbourhood The rank parameter sets the percentile value The pixels to be used are selected using a structuring element defined by the kernel parameter For more on the use of the kernel consult the documentation or help for the berode and bdilate operators When the optimization parameter is set to speed and the kernel is rectangular ie contains no zero entries a faster less precise algorithm is used For signed or unsigned byte images the results vary only where the neighbourhood extends outside ROI boundary For other image types the rank is also computed at reduced precision in the speed mode The rank operator is a generalization of other gray scale morphology operators With the rank parameter set to 50 the rank operator is equivalent to the median operator set to O it is equivalent to the the gerode operator and set to 100 it is equivalent to the gdilate operator Parameter Information kernel Type Corlmage Default 1 1 1 1 1 1 1 1 1 rank Type int Default 50 Values 0 100 optimization Type int Default 1 Values preciseness 0
352. nt the sum of differences value calculated with the upper left comer of the template image at the corresponding pixel in the input target image Values are only calculated when the template is entirely inside the target image ROI so the template can be no larger than the target ROI Demo iGraph Operators Find SoD C Prototype CorOpRtn cor_sod CorImage Target CorImage Template CorImage Sod Headers include WITHOME h wFind h Libraries WITHOME lib wObj lib WITHOME lib wFind lib 95 tmatch Template matching Description The tmatch operator performs directed template matching using sum of differences normalized cross correlation or binary comparison Given a target image and a smaller template image or a vector of template images the operator searches specified ROIs in the target for subregions which match the template image The main center output is a vector of graphic objects that specify locations in the target which matched the templates The label field of each graphic object contains its position in the output vector This label can be displayed using the overlayData operator and allows a graphic to be easily related to the corresponding element in the vector of match values The lower output is a vector of floats that represent the match value between 0 and 1 of the corresponding graphic The upper output is a vector of integers that indicates which template provided the match when a v
353. nt yOffset int blockHeight int clear int resetId int mode float aScale float bScale float aOffset float bOffset Headers include WITHOME h wWeb h Libraries WITHOME lib wObj lib WITHOME lib wWeb lib webGetBlobFeatsRoi Get Blobs and Blob features from ROIs in a stream of line scan images T F i Description The webGetBlobFeatsRoi operator performs pixel connectivity analysis within rectangular regions of interest on a stream of binary or labeled images The operator extracts run lengths from the image and outputs both a set of features describing each contiguous region and the run lengths organized into blobs in the form of a Blob vector Images in the sequence are treated as if the last line of the previous image is adjacent to the first line of the current image and the first line of the next image is adjacent to the last line of the current image allowing blobs to extend across more than one 271 image Blobs are output when they do not reach the bottom of the current image and so cannot extend into the next image in the sequence Blobs that reach the bottom row of the current image are usually saved by the operator to see if they extend into the next image in the sequence The operator can be forced to output these blobs using the reset parameter The operator is approximately equivalent to applying the getBlobFeaturesRoi operator to an image formed by concatenating the images in the st
354. o PC serial ports openSerial Open a serial port a inyin Description The openSerial operator opens a serial port for reading and writing The input object is passed through to the output unchanged The port configuration and timeouts can be set using the operator s parameters The port parameter is used to select the input port to be opened The port parameter value 0 corresponds to the COM1 port parameter value 1 corresponds to COM2 and so on Once a port has been opened the readSerial and writeSerial operators can be used to read from and write to the port An open port will remain open until it is closed by the closeSerial operator or WiT is exited Ports are not automatically closed when an igraph stops running If the specified port is already open it will be closed and reopened using the operator s parameter values When the configurePort parameter is set to Yes the serial link is configured using the baudRate parity stopBits and dataBits parameters When it is set to No the existing system settings are used Read timeouts can be specified using the rdTimeOutMult rdTimeOutConst and rdIntervalTimeOut parameters Write timeouts can be specified by setting the wrTimeOutMult and wrTimeOutConst parameters Timeout parameter values are in milliseconds No timeout is set if the parameter value is zero Parameter Information port Type int Default 1 configurePort Type int Default 0 Values No 0 Yes 1
355. oat slopeThresh int levelType float levelThresh float sampleSpacing int smoothing Headers include WITHOME h wEdges h Libraries WITHOME lib wObj lib WITHOME lib wEdges lib getXEdgesinPoly Find a edge in a polygon 76 gt eet eee adi f H Description The getXEdgesInPoly operator finds edge points in a polygon and fits a line to the points to find the best edge crossing the polygon The operator s inputs are In a grayscale image and Polygon a polygon type Graphic or Geometry object The operator searches for edges along a number of sample lines through the image that are defined by the input polygon and the numLines parameter The operator takes the best edge found on each sample line that meets the criteria specified in the operator s parameters and fits an edge line to them using a least squares fit The operator s outputs are PtEdges a vector of point type edges containing the edge points found on each sample line LineEdge a line type edge object containing the best fit line and NFound the number of edge points found The slope and level field values of the LineEdge are the average of the corresponding edge point values The direction field value is the angle the perpendicular to the line edge pointing in the direction of the rising edge It should be noted that the operator will output a line even when fewer than two edge points are found If no edge points are found the operator outpu
356. ocal thresholding using seed points Description The localThresh operator performs local thresholding for segmenting objects from a greyscale image The top input port is connected to an 8 bit input image and the bottom input port is connected to a set of seed points from which region growing is to start The seed input may be either a binary image where the seed is a non zero pixel value or a getData vector of points where each seed is an x y location The algorithm goes through the list of seeds one by one and attempts to grow the seed region by examining the seed s 8 neighbours recursively A neighbour is region grown if its pixel value satisfies a threshold The threshold is determined by the parameters peakFraction and minContrast The idea here is to grow seed values into a region with pixels greater than a certain percentage of the seed pixel value The minContrast constraint is involved to ensure the result of multiplying the seed value by the peakFraction to obtain a threshold will not go below above a minimum noise level Note how the sign of minContrast determines whether a region is considered above or below a threshold This technique for thresholding objects from a scene is ideal if the scene s background is not uniform In other words each object to be segmented has a different background level Global thresholding would either isolate too few objects or too much of the scene as a result of variation in the scene backgr
357. of templates at sparse or irregular sampling intervals The remaining inputs and outputs and the operator parameters are described in the tmatch operator documentation The bounding box produced by setting the result parameter to bounding box is the bounding box of the points specified in the templateCoor vector Parameter Information method Type int Default 1 Values Sum of Diff 0 Norm Corr 1 Binary 2 correlation Type int Default 0 Values positive 0 negative 1 both 2 threshold Type float Default 0 5 sufficient Type float threshold Default 1 numToMatch Type int Default 5 templateSampling Type int Default 1 templateXOffset Type int Default 0 templateYOffset Type int Default 0 targetSampling Type int Default 1 targetX Offset Type int Default 0 targetY Offset Type int Default 0 searchRange Type int Default 5 result Type int Default 2 Values center 0 upper left 1 bounding box 2 vicinity Type int Default 5 red Type int Default 255 Values 0 255 green Type int Default 0 Values 0 255 blue Type int Default 0 Values 0 255 Demo iGraph Operators Find TMatchS parse C Prototype CorOpRtn cor_tmatchSparse CorImage Target CorObj TemplCoor CorObj TemplVal CorObj ROI CorIntVector TemplateNumber CorGraphicVector Positions CorFloatVector CorrValue int method int correlation float threshold float sufficient_threshold int
358. of the object are determined from the const parameter based on these rules constant parameter Type Only digits no decimal point e g int32 3048 Only digits with decimal point e g _ float32 3 1416 One quoted character e g a char First character is not a digit e g String abcde 105 s with integers for x and y e g 5 Point 10 You can specify the type of the output object by setting the type parameter to the type you want This also works for compound objects too Examples parameter parameter 156 String hello hello there there I have I have quotes quotes NULL Real point 4 5 6 2 x 4 5 y 6 2 Complex 4 5e 5 real 0 00045 6 2e3 imag 6200 RGB 255 128 R 255 G 128 128 0 B 128 Other simple object types can be generated using the cast operator with the concat operator More complex objects can be created using the crObject operator Parameter Information constant Type CorString Default type Type CorString Default C Prototype CorOpRtn cor_const CorObj Const CorString constant CorString type Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib crGraphic Create a graphic object x8 amp Description The crGraphic operator constructs a new graphic object any time it receives an input The type of the new graphic is set by the graphicType parameter Gra
359. of the tiles when calculating match scores to improve the robustness of matching of partially occluded or damaged target images Tiled model images can also be used to maintain precision in 255 large greater than 256x256 model images by suppressing the automatic scaling that is applied to large model images The operator can also incorporate binary mask images into the pattern model Only pixels in the model and target images that correspond to non zero pixels in the mask image will be used in the calculation of matching scores Inputs Outputs Modell mages The input may be either a single gray scale image or a vector of gray scale images If a vector of images is used all images in the vector must be the same size and type Vectors of rotated or scaled versions of the same image can be produced by the patRotate and patScale operators These image vectors can be used as input to patInit to allow prepare a model for searches for rotated or scaled patterns in the target MaskI mages When the mask parameter is set to Yes the input may be either a single unsigned 8 bit image or a vector of unsigned 8 bit images The number of images and the size of each the images must be the same as the ModelImages When the mask parameter is set to No the input is ignored PatModel The operator produces a PatModel object that includes models of each image in the input image set The pattern model enables the pattern matching operat
360. of the window if Placement has Max size checked Property Panel alias Site Dida jane pri ee Cumi tm 5 Brims r Cal EJ Desir Firn EJ veng wij t Pa d o Li Caio aj Hh O j lg ra Size The size of the volume data set displayed as width x height x depth Depth is the number of slices in the volume Apply to Similar to the same control in the image property panel See chapter Operators Libraries in User s Manual for details Surface Controls the surface color of visible objects and the color of semi transparent materials If White is selected all visible pixels are colored white If Grayscale is selected visible pixels are colored based on the pixel value If Pseudo is selected visible pixels are colored based on the pixel value mapped to the standard pseudo color colormap used in image displays If Custom is selected visible pixels are colored based on the pixel value mapped to a custom colormap specified by the user The surface map editor pops up 71 Opacity 72 automatically when the custom color mode is first selected or can be brought up to modify the existing custom map by clicking the button The map editor will then come up In all cases surface color changes are applied to the rendered image and to the Surface ramp control On the rendered image surface colors are modified by lighting and shading to give the rendered image its 3D
361. oint pl CorPoint p2 Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib filter Filter objects by type ott i ap Description The filter operator takes input objects and outputs the same object to one of two output ports depending on whether the object type matches a set 160 of specified types given by the parameters If the input object equals any of the allowable types then it is passed to the top output equal port otherwise the it is sent to the bottom output not equal port The allowable set of types are determined by the parameters type user image and graphic The type parameter allows multiple selections if the input object type matches any of the selected types it will be output at the equal port If image or Graphic types are selected the selection is further refined using the image or graphic parameters These parameters also allow multiple selections If the object type does not appear in the type list then it can be specified by selecting the user type A list of user types is specified by entering a string for the user parameter consisting of the appropriate object name The object name may be either the object s display name the name shown in the display of an object of that type by the display operator or a probe or the programming type name Type names that contain spaces must be enclosed in quotes for example Real point The operator will
362. olor images are output in 256 grayscale or 24 bit pixel format depending on color parameter This allows the printer the freedom to use whatever halftone screen to produce the best results On low cost printers the results may not look very good but you can send the same file to a professional printer and obtain flawless images Grayscale images require much more data to be transferred to the printer and so may take more time to print particularly if the link to the printer is serial The images can be more difficult to photocopy too If draft is chosen the image is dithered with the resolution of the screen used If you are using a monochrome display the output will have exactly the same number of dots as you see on the screen show frame Enable or disable a frame around the object pagesize This information allows the object to be centered on the printed page pagesize This information allows the object to be centered on the printed page If custom is chosen the size must be specified in the customSize parameter customSize Width and height of custom paper size This is particularly useful if you want to make the object using fit parameter fit a known size orientation Landscape means longer side of paper is horizontal X axis Portrait means longer side of paper is vertical Y axis format If graphic the output is printer format General objects are printed as printer encoded formatted text An image always oc
363. ompensate perspective and lens distortions and convert image coordinates to real world coordinates calibrateCamera Perform calibration of the camera to avoid distortions Description The calibrateCamera operator computes a transformation for correcting perspective and non linear lens distortions using a checkerboard pattern The resultant calibration object can be used by the restoreImage and restoreCoordinates operators to restore images or coordinates graphic objects respectively To use this operator a physical checkerboard should be prepared and imaged The checkerboard must consist of high contrast and uniform black and white squares When the image is taken the checkboard must fill the entire image area It is all right to have squares only partially visible around the edges of the image The length of the side of each square can be set by the cellSize parameter The interpolation parameter determines whether perspective or bilinear interpolation is used within each square The origin parameter can be used to set the origin and direction of the axes of the coordinate system of the transformation If set to none then coordinate system is computed to maximize the amount of the image visible after restoration with the Y axis pointing down X axis pointing right and origin close to the left top corner of the image In all other cases the origin set at the corner of the square closest to the selected corner and have both edge
364. onding output vector element is the empty string Parameters 229 noise The noise parameter specifies the minimum low to high or high to low change in pixel value necessary to constitute the edge of bar type The type parameter specifies the type of Pharmacode to be decoded Parameter Information noise Type int Default 5 type Type int Default 0 Values standard 0 wide space 1 complementary 2 Demo iGraph Operators Matrix Pharmacode igr C Prototype CorOpRtn cor_getPharmacode CorImage Target CorObj ROI CorObj Decoded int noise int type Headers include WITHOME h wMatrix h Libraries WITHOME lib wObj lib WITHOME lib wMatrix lib getRSS14 Finds and decodes RSS 14 bar code sii 230 Description The getRSS14 operator accepts a binary or grayscale image and one or more rectangular regions of interest ROIs then searches the ROIs for the RSS 14 barcodes If a Composite Component is found in a region it is decoded to a string too Inputs Target The input image must be an unsigned 8 bit grayscale image with the barcode represented as dark bars on a light background Roi A single ROI may be specified by a rectangular Graphic or Geometry object Multiple ROIs can be specified by a vector of Graphic or Geometry objects Outputs Decoded If the ROI is a single object then the output is a String object If a valid RSS14 is found the output string contains the
365. oordinate of a pixel on the blob perimeter lower right comer perimy Y coordinate of a pixel on the blob perimeter lower right comer xc X coordinate of blob centroid sumx n ye Y coordinate of blob centroid sumy n angle Angle of best fit ellipse major axis clockwise from x axis major Length of best fit ellipse major axis minor Length of best fit ellipse minor axis perim Perimeter data for blob must run the getPerimeter operator to fill in this structure Blobs The Blobs output is a vector of Blob objects Each Blob object contains a single run length The run lengths are grouped into contiguous regions Within each group the run lengths are sorted by increasing line number and then by the x location of the end of the run The groups are sorted on the last run length in the group The operator keeps a running line count the run lengths and information derived from them are referenced to the beginning of the first image in the sequence of images Parameters connected The connectivity of the output blobs is specified by the connected parameter When this parameter is set to 4 pixels that share a common edge are considered to be connected so each pixel has four neighbours When it is set to 8 pixels that touch corner to corner are also considered connected so each pixel has eight neighbours type The type parameter determines how the input image will be interpreted When it is set to binary
366. opWatch Time igraph operations a W F b Description The stopWatch operator controls a global timer which can be used for timing igraph operation The object received at the input initiates the stopWatch action and is passed through to the Done output port unchanged The stopWatch action is controlled by the reset parameter When the parameter is set to Reset the timer is restarted and 0 is output from the Time output port When the reset parameter is set to Read the time in milliseconds since the last Reset is produced on the time output port and the timer continues All stopWatch operators in an igraph affect the same global timer Parameter Information reset Type int Default 1 Values Reset 0 Read 1 wrText Write object as a text string Description The wrText operator writes an object to a file in a text format The text can be written in a labelled or unlabelled format as selected by the format parameter 213 For non image data the labelled text is written to the file in the same format as that produced by the display operator For images only header information giving the image size and type is written In the unlabelled mode the operator converts a simple WiT object such as a number or string or a vector with a simple object base type to a text string and writes the string to disk Attempting to print images or complex data types in the unlabelled mode will result in an error String objects are
367. operator will offset graphics or other geometric objects by an amount specified in the input graphic This operator is primarily used to return geometric objects found in regions of interest extracted from an image using the extract operator to the original image coordinate system The Locations input receives a rectangle or point type graphic or vector of graphics that specifies the location of the origin of the coordinate system of the geometric objects received at the In input port If a single graphic is received at the Locations input the geometric object may be a Graphic Geom Line Edge Point or Fpoint object or a vector of one of these types If a vector is received at the Locations input a vector of the same size must be received at the In port The elements of this vector may be single geometric objects or vectors of geometric objects as described above The elements of the Locations input vector are applied to the corresponding elements of the In input vector Demo iGraph Operators Fit warp igr C Prototype CorOpRtn cor_insertGraphic CorObj In CorObj Locations CorObj Out Headers include WITHOME h wGeom h 126 Libraries WITHOME lib wObj lib WITHOME lib wGeom lib matchSize Make two images the same size by clipping or padding as a ha Pi Description The matchSize operator takes two input images A and B offsets one image relative to the other and pads or clips one or both
368. or Property Panel z Srl propertie Aran I x o SigevT pre PGS ri m Bange 2 1128 70 Areata Coe E Ecler bor irava e af a t 4 ae A Ai Gai wie code Loa Size Type Original width and height of image and the type of each pixel ge Range Similar to the same control in the image property panel See chapter Operators Libraries in User s Manual for details Frame Similar to the same control in the size image property panel Apply to Similar to the same control in the image property panel 68 Color Similar to the same control in the Map image property panel Viewing Controls the viewpoint from which the surface is observed Image Similar to the same control in the index image property panel Grid Brings up the surface grid panel see below for adjusting the grid and wireframe used to render the surface Wire color Brings up a color panel for changing the color used to draw the wireframe and axis labels Load all _ Similar to the same control in the image property panel Grid Panel sef Bwege Wrerang 16 Wix The magnification applied to the image intensity to compute the height of each point on the surface A smaller magnification means the surface will appear flatter Pixel per square The 3 D surface is drawn as a rectangular mesh Each rectangle square on the surface represents an area of o
369. or degraded versions of that vector s character MaskI mages When the mask parameter is set to Yes the input must be a vector of image vectors in which the images are unsigned 8 bit images The number of images and the size of each the images must be the same as the corresponding ModelImages When the mask parameter is set to No the input is ignored PatModels The operator produces a vector of PatModel objects Each element of the vector includes models for the images in one of the image vector elements of the input vector of vectors Parameters 258 subPixel If the subPixel parameter is set to Yes coefficients required for subpixel pattern matching will be calculated and included in the pattern model These coefficients must be precalculated by the patInit2D operator in order to use the pattern models for subpixel resolution matching mask If the mask parameter is set to Yes the MaskImages input will be used to create a masked pattern model in which only pixels corresponding to non zero pixels in the mask image will be used to calculate the match score If it is set to No all pixels in the model image and corresponding region of the target image will be used in the score calculation tileCols and tileRows If the tileCols and tileRows parameters are used to set the number of tiles into which the model image is divided horizontally and vertically The model image is not tiled when both parameters are set to 1
370. or get pixel values at points rows columns arbitrary lines and rectangular regions of images extract Extract a region of an image 5 She Description The extract operator extracts a rectangular subimage or images from the input image Subimages are specified by the subImage parameter If this parameter is a single Graphic or Geometry object an image or a blob vector a single subimage is produced If the parameter is a vector of Graphic or Geometry objects or a vector of images a vector of subimages is produced For a Graphic or Geometry object the subimage produced will be the smallest rectangular region enclosing the object For an image the subimage will correspond to the image ROI The subImage parameter is often supplied as an input parameter Suitable input objects can be produced by operators such as createRect which creates a single rectangle or getData which produces a vector of Graphics The subimages can be inserted back into the image after processing using the insert operator Geometric objects such as Graphics Edges Lines and Points can be returned to the orginal image coordinates using the insertGraphic operator Parameter Information subImage Type CorObj Default OBJ_B T CorGeom 4 CorFpoint 2 0 0 63 63 C Prototype CorOpRtn cor_extract_1 CorImage In CorObj Out CorObj subImage Headers include WITHOME h wPixel h Libraries WITHOME lib wObj lib WITHOME lib wPixe
371. or_getMatch CorImage In CorObj ROI CorGraphicVector Positions int numToMatch int selectMatches int vicinity int searchRange float threshold int find int color int red int green int blue 93 Headers include WITHOME h wFind h Libraries WITHOME lib wObj lib WITHOME lib wFind lib getPeaks Find local maxima or minima Description The getPeaks operator finds local maxima or minima in a grayscale image For this operator a local maximum minimum is defined as a single pixel that whose value is greater less than the values of the eight connected pixels or a group of up to four pixels in a 2x2 block whose values are greater less than the values of the 12 pixels surrounding the 2x2 block The operator output is a vector of Point value objects in which the value field is the pixel value at the peak and the pt field is the peak location The peaks are sorted in decreasing order of value for maxima and in increasing order for minima The type parameter is used to specify whether the operator should search for Minima or Maxima The threshold and relThreshold parameters are used to set a threshold for the peak pixel values Only peak pixel values above the threshold for maxima or below the threshold for minima are accepted The relThreshold can be used to set the threshold to a percentage of the actual range of values in the image For maxima the relative threshold value is set to the specified
372. order as specified by the direction parameter The output parameter is normally set to object which produces an output vector containing the the objects from the input vector in sorted order The output parameter may be set to index in which case the output is an integer vector where each element is an index into the input vector for the object rather than the object itself This is most often used with vectors of simple numeric objects Typical examples of applications of this operator include sorting a vector of Point or FPoint objects on their x or y fields sorting Feature vectors by size of the feature using the npixels or boxarea fields or sorting a vector of images on their height H field or width W field Parameter Information field Type CorString Default x direction Type int Default 0 Values ascending 0 descending 1 output Type int Default 0 Values object 0 index 1 C Prototype CorOpRtn cor_sortObj CorVector InVector CorVector SortedVector CorString field int direction int output Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib 171 sortTwoFields Sorts a vector of complex objects based on the values of two fields s Description The sortTwoFields operator allows the user to sort a vector of complex objects based on the values of two fields in the objects The objects are sorted on the first field and then
373. orresponding to different methods of averaging the neighbouring pixels are selected using the are determined using the type parameter The filter weights for each type are Type 0 0 Z 4 Pa Oo 0 Type 1 x g a 8 set Type 2 2 lt 2 4 2 2 This operator could be executed in two modes The user may choose the mode by using the optimization parameter If the parameter value is speed then execution goes way faster otherwise operator delivers more precise results The operator accepts integer and float type input images and color images in the case optimisation parameter value is speed An unsigned integer input image produces a signed integer output image with the same number of bits as the input image In all other cases the output image is the same type as the input image Parameter Information type Type int Default 0 Values 0 0 1 1 2 2 optimization Type int Default 1 Values preciseness 0 speed 1 Demo iGraph Operators Filter Entropy Operators Filter Variance C Prototype CorOpRtn cor_laplace CorImage In CorImage Out int type int optimization CorOpRtn cor_laplace_consume CorImage In int type int optimization Headers include WITHOME h wFilter h Libraries WITHOME lib wObj lib WITHOME lib wFilter lib line Directional line enhancement Description The line operator enhances lines with a specific orientation in the image The line
374. ors Blobs AndBlobs igr Operators Blobs Holes igr C Prototype CorOpRtn cor_AndBlobs CorBlobVector A CorBlobVector B CorBlobVector And CorBlobVector NAnd Headers include WITHOME h wBlob h Libraries WITHOME lib wObj lib WITHOME lib wBlob lib getBlobFeatures Get Blobs and Blob features from a binary image PELi cm Description The getBlobFeatures operator computes a number of features of blobs connected regions in a binary or labeled image The operator also produces the blobs in a run length format The blob perimeters are optionally calculated The input image must be an unsigned 8 or 16 bit binary or labeled image A Blob feature vector describes an object by measuring centroid location major and minor axes bounding boxes etc The Blob feature vector is sent to the output for use by analysis operators for object recognition Blob features may be viewed using a combination of the plotFeatures and overlayData operators The Blob feature vector consists of the following fields Field Description Name id Unique blob identifier npixels Number of pixels in the blob area if pixels are square xmin Minimum x coordinate of blob left edge xmax Maximum x coordinate of blob right edge ymin Minimum y coordinate of blob top edge ymax Maximum y coordinate of blob bottom edge sumx Sum of all x coordinates sumy Sum of all y coordinates sumxsq Sum of all x coordina
375. ors to quickly search the image for matches to the patterns represented by the input image set Parameters 256 subPixel If the subPixel parameter is set to Yes coefficients required for subpixel pattern matching will be calculated and included in the pattern model These coefficients must be precalculated by the patInit operator in order to use the pattern models for subpixel resolution matching mask If the mask parameter is set to Yes the MaskImages input will be used to create a masked pattern model in which only pixels corresponding to non zero pixels in the mask image will be used to calculate the match score If it is set to No all pixels in the model image and corresponding region of the target image will be used in the score calculation tileCols and tileRows If the tileCols and tileRows parameters are used to set the number of tiles into which the model image is divided horizontally and vertically The model image is not tiled when both parameters are set to 1 The image will be broken into the equal size tiles If the number of pixels in the model image does not divide evenly by the number of tiles remaining pixels on the right and or bottom of the model image will be dropped colorMode The colorMode parameter controls the normalization of color images When this parameter is set to brightness the model image and target image are converted to gray scale images corresponding to the Y channel of the
376. ototype CorOpRtn cor_getBarcode CorImage Target CorObj ROI CorObj Decoded CorObj Type CorGraphicVector ScanLines int minSize int maxSize int invert int noiseLev int threshold int barcodes int checkSum int timeout Headers include WITHOME h wMatrix h Libraries WITHOME lib wObj lib WITHOME lib wMatrix lib getCCA Finds and decodes Composite Component A of the RSS 14 bar code ME o Description The getCCA operator accepts a binary or grayscale image and one or more rectangular regions of interest ROIs then searches the ROIs for 221 Composite Component A of the RSS 14 barcodes If a Composite Component is found in a region it is decoded to a string Inputs Outputs Target The input image must be an unsigned 8 bit grayscale image with the barcode represented as dark bars on a light background Roi A single ROI may be specified by a rectangular Graphic or Geometry object Multiple ROIs can be specified by a vector of Graphic or Geometry objects Decoded If the ROI is a single object then the output is a String object If a valid Composite Component is found the output string contains the characters represented by the Composite Component If no valid Composite Component is found an empty string is output If the ROI is a vector of rectangles then the operator seeks for a Composite Component in each ROI The output is a vector of String objects If a valid Co
377. ound This operator is often used after high pass filtering an image where there is a definite base level with which to measure contrast The seed points can be generated by performing an edge detection followed by zero crossing detection and then using the zero crossing points as seeds Parameter Information peakFraction Type float Default 50 minContrast Type int Default 0 Demo iGraph Operators Segmentation LocalThreshold C Prototype CorOpRtn cor_localThresh CorImage In CorObj Seed CorImage Out float peakFraction int minContrast Headers include WITHOME h wSegment h Libraries WITHOME lib wObj lib WITHOME lib wSegment lib thresh Threshold an image in an interval Description The thresh operator takes a greyscale or color image and produces a binary or semi binary image based on a specified threshold range For grayscale images the thresholds are applied directly to the values of image pixels For color images the thresholds are applied to a luminence or intensity value for each pixel the Y channel for Yuv images the V channel for HSV images and a luminence value derived from all three channels equivalent to the Y value for RGB images In the simplest use set invert to No and levels to single Then pixels which are equal to or above lo are turned on and those below are turned off The hi parameter is not used when levels is set to single The value of on and off i
378. ow the x and y fields in Real Point objects are converted to the corresponding fields in the Point object Selecting round will set the output value to the nearest integer of the input value floor will set it to the nearest integer less than or equal to the input value and ceiling will set it to the nearest integer greater than or equal to the input value Casting to a String or File type from one of the floating point integer or point including Real Point types results in a text string representation of the value For example an int object with value 10 will be converted to the string JO and a Point object with x 0 y 0 will be converted to the string 0 0 Input String or File type objects can be cast to numeric types When a string is converted to a signed or unsigned integer type it is interpreted using the C programming language convention Strings starting with a 0 are interpreted as octal numbers and strings starting with Ox or OX are interpreted as hexadecimal numbers For example the strings 12 014 and OxC are all cast to the integer 12 If the entire string cannot be converted to the specified numeric type an error occurs halting the igraph Parameter Information type Type int Default 0 Values byte 0 ubyte 1 short 2 ushort 3 int 4 uint 5 float 6 double 7 String 8 File 9 Point 10 Real point 11 char 12 hex8 13 hex16 14 hex 15 conversion Type int Default 0 Val
379. pRtn cor_gaussPyramid CorImage In CorImageVector Pyramid int filter int minsize 187 Headers include WITHOME h wPyramid h Libraries WITHOME lib wObj lib WITHOME lib wPyramid lib reduce Filter and decimate an image o r Description The reduce operator applies a filter to an input image and then decimates the resulting image by eliminating every other row and column The filter parameter denotes which gaussian filter to use The filters are discussed in more detail in the gauss operator within the filter library The None option for filter indicates that no filtering will occur and that the input image will simply be decimated Parameter Information filter Type int Default 0 Values None 0 2x2 1 3x3 2 5x5 3 7x7 4 C Prototype CorOpRtn cor_reduce CorImage In CorImage Out int filter Headers include WITHOME h wPyramid h 188 Libraries WITHOME lib wObj lib WITHOME lib wPyramid lib showPyramid Display all the images in a pyramid as one image Ft Description The showPyramid operator takes an input image vector and creates a single image containing all elements within the vector Component images can be ordered either horizontally vertically or in a spiral fashion as indicated by the orientation parameter If a positive integer value J is specified for the spacing parameter then a zero valued band of pixels of width J will be placed bet
380. parameter should be set to the same value for both transformations The width and height parameters set the width and height of the rectangular coordinate space output image Parameter Information direction Type int Default 0 Values R gt P 0 P gt R 1 rBins Type int Default 128 minRadius Type float Default 0 maxRadius Type float Default 127 aBins Type int Default 360 minAngle Type float Default 0 maxAngle Type float Default 359 width Type int Default 256 height Type int Default 256 type Type int Default 0 Values Polar 0 Log Polar 1 Demo iGraph Operators Fit polar igr C Prototype CorOpRtn cor_polar CorImage In CorObj Centroid CorImage Out int direction int rBins float minRadius float maxRadius int aBins float minAngle float maxAngle int width int height int type Headers include WITHOME h wGeom h Libraries WITHOME lib wObj lib WITHOME lib wGeom lib rotate Rotate an object by an arbitrary angle ay Description The rotate operator performs an arbitrary rotation on its input object The angle parameter defines the 129 angle in degrees where positive angles represent a rotation counter clockwise The mode parameter can be used to specify the size of the output image In the expand mode the operator will produce an output image at least large enough to accomodate the entire rotated image with the center pixel of the ro
381. pe int Demo iGraph Operators Edges X EdgesOnLine C Prototype CorOpRtn cor_traceEdges CorImage EdgeImage CorGeomVector Edges int Truncated int type float threshold int minLength int mode int maxMemSize Headers include WITHOME h wEdges h Libraries WITHOME lib wObj lib WITHOME lib wEdges lib Filter The Filter library supports convolution filtering of images The library includes general convolution operators as well as optimized special cases of the convolution filters and operators which produce images derived from convolved images compass 8 point directional edge enhancement Description The compass operator enhances edges in the input image perpendicular to one of eight directions Edge enhancement is often a preliminary step to edge finding and segmentation procedures The operator applies a first order derivative based filter to the input image The orientation of the derivative is specified by the direction parameter and can be set to one of the eight compass points N NE E SE S SW W or NW where N is north or up in the image negative y direction and E is east or right positive x direction and so on The compass direction selects filter weights such that the magnitude of the values in the output image will be largest for edges perpendicular to the compass direction and will fall to zero for edges parallel to the direction Rising edges produce positive values and fall
382. pe int Default 0 spacing Type int Default 0 location Type int Default 0 Values top left 0 center 1 C Prototype CorOpRtn cor_tileImages CorImageVector In CorImage Out int tileColumns int tileRows int spacing int location Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib 173 vecRedim Change the dimensions of a vector but retain element values Description The vecRedim operator changes the dimensions of a vector or image object while preserving all the element values The number of elements in the input object must match the number of elements specified If the height parameter is 0 then the output is a vector If height is 1 or greater the output is an image If the autoSet parameter is set to width height must be non zero and the number of elements in the input object must be divisible by height Similarly for width Parameter Information autoSet Type int Default 0 Values none 0 width 1 height 2 width Type int Default 1 height Type int Default 0 C Prototype CorOpRtn cor_vecRedim CorObj In CorObj Out int autoSet int width int height Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib 174 Pixel Access The Pixel Access library supports reading and writing pixel values from or to images The library includes operators that set
383. pe values For scalar inputs this is a float vector For point inputs this is a real point vector that contains the slope values in the y field The x field contains the x values from the Profile input The interpolate parameter can be used to specify the use of interpolation to more accurately determine the position of the maximum slope location This is only applied to the location field the x field in the MaxSlope output The maximum slope value and the position and value at the start and end of the slope are not interpolated Parameter Information direction Type int Default 0 Values rising 0 falling 1 rising or falling 2 threshold Type float interpolate Type int Default 1 Values No 0 Yes 1 Demo iGraph Operators Statistics LocateInProj_Slope C Prototype CorOpRtn cor_getSlopeMaxima CorVector Profile CorVector Slope CorFpointVector SlopeMax CorFpointVector SlopeStart CorFpointVector SlopeEnd int direction float threshold int interpolate Headers include WITHOME h wStats h Libraries WITHOME lib wObj lib WITHOME lib wStats lib histCentroid Compute the centroid of a histogram or numeric vector Lie Description The histCentroid operator computes the centroid of a histogram or vector of numbers Histograms are implemented as vectors of Real point objects where the x field represents the value and the y field represents the frequency The centroid of the
384. pecified for removal Outputs Out The output image contains the rows of the input image that remain after the specified number of rows are removed from the top and bottom Parameters rows The rows parameter specifies the number of rows to be removed from the top and bottom of the image Parameter Information rows Type int Default 2 C Prototype CorOpRtn cor_webRowRemove CorImage In CorImage Out int rows Headers include WITHOME h wWeb h Libraries WITHOME lib wObj lib WITHOME lib wWeb lib webScrollGraphics Subtract from one or a vector of Graphics Description The webScrollGraphics operator takes a Graphic or vector of Graphics and subtracts the specified offset value from the y coordinates of all points in the graphic This operator is typically used to transform graphics produced using the output of the webGetBlobs operator into the frame of reference of the images produced by the webWaterfall operator In this case the offset value is taken from the o1 output of the webWaterfall operator The transformed graphics can then used with the waterfall image for overlay display or extraction of regions of interest It can also be used to convert the graphics to the frame of reference of images from the original image stream by feeding the image stream into the webLineCounter operator and using its yOffset output to control the offset Inputs InGraphics The input is a Graphic or vecto
385. perator or vectors of scale images such as those produced by the patScale operator output The Candidates setting of of the output parameter allows the user to output the locations match scores and pattern indexes of the candidates rather than those of the refined matches normally produced The Candidates setting is usually used during development to help determine the best settings for the various parameters that control candidate selection The MultipleMatches Sub Panel The MultipleMatches sub panel includes parameters that control the output when multiple matches are required max Matches The maxMatches parameter sets the maximum number of matches that will be produced by the operator The operator may produce fewer matches relThresh The relThresh parameter is used to set a relative rejection threshold for the matches The relative rejection threshold is determined by multiplying the parameter value by the match score of the best match found This value is then applied like the normal rejection threshold xVicinity and yVicinity These parameters specify a minimum region around a match location in which matches with lower match scores will be suppressed Two matches will never be produced that are less than both xVicinity pixels apart in x and yVicinity pixels apart in y Parameter Information rejectThresh Type float Default 0 acceptThresh Type float Default 1 minContrast Type float Default
386. phic objects require a point list which identify the location of the object and in some cases other parameters of the object e g radius of a circle The point list for a graphic object is entered by selecting the Edit panel for the points parameter This panel presents an array editor where each row in the array represents a point The first column is the x coordinate and the second is the y coordinate A single point is required to specify a Point location A single point is also used to designate the starting position for Text Two endpoints are required for a Line Polyline and Polygon types use a variable length list of points to designate vertices where at least one point is required A Rectangle is specified by two points representing the upper left and lower right corners of the rectangle A Circle is specified by two points with the first point the center of the circle and the x coordinate of the second point the radius The y coordinate of the second point is not used An Arc object represents an elliptical arc It requires four points in its point list The first second and fourth points specify the underlying ellipse or circle for the arc The third point defines the part of the ellipse covered by the arc First point in the point list represents the center of the ellipse The x coordinate of the second point is half the length of the major axis of the ellipse typically called a and the y coordinate is half the length of
387. pisal Length 100 314 Beale E i ee T Blot perimeter Heig 103 Depi E Comber Aex 0 i Sapaata Waar S0 48502 i Replace Sbd Dew 41 317 Mirt 49 Mae 355 Posio 51 111 Type Type of graphic being used for measuring Length Length of graphic Width Width of bounding box of graphic Height Height of bounding box of graphic Area Area of graphic Std Dey Standard deviation of pixel values within graphic area Min Minimum pixel value within graphic area Max Maximum pixel value within graphic area Positions List of x y coordinates of each vertex of graphic Units Unit name From operator UnitName parameter Scale Ratio of pixel count to physical length For example if Scale is set to 2 and a line 5 pixels long is entered then a length of 10 physical units will be reported The initial value of Scale is set from the UnitsPerPixel operator parameter You can change it interactively by clicking the button Plot perimeter If checked the pixel values along the perimeter of the graphic are plotted whenever a new graphic is entered Does not apply to point or text graphic types You can control whether plots should be done in one single graph or as separate graphs with the Graph control below Graph Control the placement of new perimeter plots If Combine is chosen subsequent plots are merged with the existing graph This makes it easy to compare
388. pixel values in the pattern model image This parameter is usually set to a value between zero and one although values greater than one can be used if the target image has higher contrast than the model images maxTilesOmitted If the pattern model breaks the pattern images into multiple tiles the maxTilesOmitted parameter allows the operator to omit some of the tiles while calculating the best matching score The operator will use the best score for a location that omits at most the number of tiles specified by this parameter The operator will always use at least one tile to calculate the best score Setting this parameter to a value greater than zero provides some tolerance to partially occluded or damaged target regions localSearch The localSearch parameter specifies the method used to search the local neighbourhoods of candidates found in the low resolution search When the parameter is set to Fast a binary hill climbing method is used The hill climbing method can get stuck in local a maximum in particular when the target and pattern model images contain fine regular patterns To prevent this occuring the parameter can be set to All in which case an exhaustive search of the neighbourhood is performed The exhaustive search is slower for large candidate search scale factors subPixel The subPixel parameter allows the match locations to be refined to subpixel precision The pattern model must have been prepared using the
389. plicate of the input vector with the blob perimeter data set Various statistics are generated as part of each element in the perimeter structure For example the maximum and minimum distance between each blob centroid and all of its perimeter pixels is provided As with the getBlobs operator any non zero pixel is considered to be contained in a blob The perimeter fields are defined as follows perimLength Total length of perimeter of points around object maxDeltaX Max distance along x axis from x centroid minDeltaX Min distance along x axis from x centroid maxDeltaY Max distance along y axis from y centroid minDeltaY Min distance along y axis from y centroid sumRadius Sum of radial distances of each perimeter point from centroid sumRadiusSqrd Sum of radial distances squared a Max radial distance a minRadiusSqrd Min radial distance Min radial distance squared maxRadDeltaX Max distance along x axis from x centroid at max radial point maxRadDeltaY Max distance along y axis from y centroid at max radial point minRadDeltaX Min distance along x axis from x centroid at max radial point minRadDeltaY Min distance along y axis from y centroid at max radial a maxLength Max Max length of perimeter of perimeter minLength Min length of perimeter mawaa Max width of Mansa minWidth Min width of Min width of perimeter points Point vector containing all perimeter points Th
390. put object because the kernel when aligned with any input pixel will set the output only if both the pixel corresponding to the kernel center is set and if the pixel one to the right of that input pixel is set In a similar fashion the second example reduces the object on the left side The third example uses a kernel of size 2x2 with its center being the top left bit in the kernel The resulting output object has been reduced to only two pixels because only two set pixels in the input object have neighbours below and to the right which are both set All other pixels are cleared The implementation of erosion used in WiT is defined in the paper by Haralick Sternberg and Zhuang IEEE Transactions on Pattern Analysis and Machine Intelligence July 1987 Border conditions are handled by assuming all pixels outside of the input image have non zero value Note that when 144 the center of the kernel is zero it is possible for erosion to set an output pixel even if the input pixel is zero Greyscale erosion is performed by computing the minimum value within every neighbourhood of the input image As with binary erosion the neighbourhhod may be either a rectangular region centred on the pixel or arbitrary pixels specified by the kernel parameter array Pixels beyond the image boundary are assumed to be the maximum value possible for the type of image being processed This means that they will have no effect when dealing with a minimum cal
391. put object is an image then edit brings up a simple painting tool which allows you to modify the pixel values in the image It displays the image magnified according to the magnification parameter The higher the magnification the larger each pixel will appear Edit also displays a property window next to the image xi Bushee TP grat We a 2 A ll Hi Bersi When edit is running any left mouse buttons clicks or mouse drags inside the image will set the pixel s to the pixel value indicated by the Pixel value control in the edit properties window Brush size controls the area painted for each mouse position Brush sizes are always square in shape The button switches the operation mode to sample mode In sample mode the cursor changes to When you click the left mouse button the value of the pixel is copied to the Pixel value control in the edit property window After one sample edit automatically switches back to edit mode The x button terminates the edit session and sends the modified image to the output The E button terminates the session but does not produce any output The modified image is discarded The Reset button resets the image back to the original If the input type is not an image then edit brings up the Object Editor See the User s Manual for details about the Object Editor For WiT Professional Toolkit users the edit operator supports the witControlDisplay function for the commands brush
392. r If the parameter value is speed then execution goes way faster otherwise operator delivers more precise results Binary or gray scale dilation is selected using the type parameter The input for binary dilation must be an 8 bit unsigned image The result is a binary 8 bit unsigned image For gray scale dilation the input may be any integer or float gray scale image or color in the case optimisation is speed The output image is a gray scale image of the same type as the input image The kernelType parameter is used to select the type of structuring element or neighbourhood specifying kernel used The structuring element may be specified as a rectangular region with the size of the region is given by the width and height parameters or it may be specified as an arbitrary 141 region in which case the neighbourhood is specified using the kernel parameter Dilation by a rectangular structuring element is usually much faster than dilation by an arbitrary element particularly for larger neighbourhoods One common exception is dilation by 3x3 structuring elements where the arbitrary kernel usually is faster Binary dilation tends to make objects larger connecting discontinous objects and filling in holes The kernel either rectangular or arbitrary specifies the structuring element which will control the dilation The center of the kernel is always in row height 2 and column width 2 For each pixel p in the input image if any pixe
393. r and the graphic objects are superimposed The rangeType parameter allows the user to map the image minimum and maximum by selecting Actual the image type range minimum and maximum by selecting Max or a user specified minimum and maximum by selecting User The Max range setting is not valid for floating point images the actual image range values are used Values outside the user specified range are clipped to the range when the User option is selected The graphical objects can be superimposed in grayscale or color as specified by the outType parameter The colors associated with the graphical object are specified on the Color subpanel The source of the graphical object color is determined by the useColor parameter If the parameter is set to original the colors are taken from the the color fields of the graphical object A luminance value determined from the red green and blue channel values is used for grayscale graphics in the original color mode If useColor is set to forced the colors are specified using the operator s parameters The outline and fill parameters are used to specify drawing the outline and filling the graphical object Both parameters may be on in which case the outline is superimposed on the filled object The pen color for object outlines and the fill color are specified using the penval and pixval parameters respectively for grayscale objects and using penR penG penB and fillR fillG fillB for color The
394. r which is output on the upper port and those that do not are collected in the FalseVector which is output on the lower port The output parameter is normally set to object which produces output vectors containing the the objects from the input vector The output parameter may be set to index in which case the output vectors are integer vectors where each element is an index into the input vector rather than the object itself 165 This is most often used with vectors of simple numeric objects Parameter Information field Type CorString Default x value Type float Default 30 hiValue Type float Default 0 condition Type int Default 2 Values 0 gt 1 lt 2 3 gt 4 lt 5 range 6 output Type int Default 0 Values object 0 index 1 C Prototype CorOpRtn cor_partition CorVector InVector CorVector TrueVector CorVector FalseVector CorString field float value float hiValue int condition int output Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib pf Insert a new field value into input object oh 166 Description pf performs the same function as putField pf has a small icon which makes it more convenient to use in large igraphs See Also putField Parameter Information field Type CorString Default x C Prototype CorOpRtn cor_putField CorObj In CorObj FieldValue CorObj
395. r contain the fields p and p2 the two endpoints of the line length the line length and angle the angle between the line and the positive x axis in degrees The mergeLines operator can be used to combine approximately collinear line segments with small gaps between them into longer lines Parameter Information deviation Type int Default 2 Demo iGraph Operators Geometry convexHull igr C Prototype CorOpRtn cor_geom21lines CorGeomVector In CorLineVector Out int deviation Headers include WITHOME h wFit h Libraries WITHOME lib wObj lib WITHOME lib wFit lib 105 getBoundingBox Find bounding box of points graphic or image i H F Description The getBoundingBox operator finds the bounding box of an image a set of points a Graphic or Geometry object or of a vector of Graphic or Geometry objects The operator s output is a rectangle type Graphic object For image inputs the bounding box is the minimum rectangle containing all non zero pixels in the image ROI The input image may be any image type The operator will accept vectors of Point Real Point Point Value Graphic Geometry Edge or Line objects or single Graphic Geometry Edge or Line objects The output is the minimum rectangle that contains the objects In the case of text Graphic or Geometric objects only the text reference position the lower left corner of the text string is used to calculate the bounding box De
396. r executes device dependent setup for controlling subsequent acquisitions The following applies only to servers that support the CAB Coreco Auxilliary Bus acqSetup commits hardware resources that are required for single frame acquisition mode Normally it is not necessary to execute this operator explicitly since it is called implicitly by the acquire operator if required For acquisition to the CAB it is necessary to execute acqSetup once after all cabSetup operators have executed but before any cabStart operators have executed See the help for any of the CAB operators for more information acqSetup may also be used to ensure that the first execution of acquire not incur the overhead necessary to set up the hardware This operator should not be used if acqStart is also executed C Prototype CorOpRtn acqSetup Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqSize Set input window on frame grabber 300 Description The acqSize operator is used to set a region of interest for transfer from a frame grabber to host The frame grabber must support a hardware ROI transfer capability This operator can be used to transfer a limited region and thereby increase transfer performance The w h x and y parameters are used to set the width height and origin offset of the region of interest relative to the camera image Parameter Information x Type int Def
397. r of Graphics Outputs ScrolledGraphics The output is the input Graphic or Graphic vector with the offset value subtracted from the y value of each point Parameters offset The offset parameter specifies the amount subtracted from the y coordinates of all input graphics When this operator is used with other web operators this is usually an input parameter Parameter Information offset Type int Default 0 C Prototype CorOpRtn cor_webScrollGraphics CorObj InGraphics CorObj ScrolledGraphics int offset Headers include WITHOME h wWeb h Libraries WITHOME lib wObj lib WITHOME lib wWeb lib webWaterfall Create a waterfall image from a stream of images m 287 Description The webWaterfall operator appends the input image to the bottom of an image produced by appending previously received images and outputs a specified number of rows from the bottom of the new image If the input images are vertically adjacent blocks from a single imaged space the webWaterfall output images can be viewed as a moving window into the image space Inputs ImageBlock The stream of images input to an instance of the webWaterfall operator must all be of the same size and type The input image height must be less than the height specified for the output image Outputs Waterfall The output images are of the same width and type as the input images The output image height is specified by a parameter The
398. r of String objects If a valid barcode is found in a ROI the corresponding element of the output vector contains the decoded string If no valid barcode is found in a ROI the corresponding output vector element is the empty string Parameters noise The noise parameter specifies the minimum low to high or high to low change in pixel value necessary to constitute the edge of bar Parameter Information noise Type int Default 5 Demo iGraph Operators Matrix EAN 8 igr C Prototype CorOpRtn cor_getEAN8 CorImage Target CorObj ROI CorObj Decoded 226 int noise Headers include WITHOME h wMatrix h Libraries WITHOME lib wObj lib WITHOME lib wMatrix lib getECC200 Finds and decodes ECC 200 2D barcodes a HH k Description The getECC200 operator accepts a binary image and searches for ECC 200 barcodes If a barcode is found then it is decoded to a string If several codes are found then several strings are generated as output Inputs Target The input image must be an unsigned 8 bit binary image with the barcode represented as white bars or dots on a black background The ECC 200 codes may be rotated scaled or warped Outputs Decoded The Decoded output is a vector of String objects If a valid barcode is found the output vector contains a string with the characters represented by the barcode If no valid barcode is found an empty vector is output Info The Info output provide
399. rImage In int filterSize int optimization Headers include WITHOME h wFilter h Libraries WITHOME lib wObj lib WITHOME lib wFilter lib grad Gradient magnitude and direction filter Description The grad operator computes an approximation to the gradient of the input image The gradient kernel is selected by the type parameter to use either Sobel or Prewitt filter weights The operator accepts integer or float input images The operator produces two output images representing the magnitude and direction of the gradient vectors at each pixel The direction is in degrees in the range 180 to 180 in the case degrees is 360 or in special units in the range 128 to 127 if degrees is 256 The magnitude output is the same type as imput image type and orientation is float if parameter degrees is 360 and 8 bit integer otherwise The top output is the magnitude image and the bottom output is the direction image of the gradient vectors This operator could be executed in two modes The user may choose the mode by using the optimization parameter If the parameter value is speed then execution goes way faster otherwise operator delivers more precise results In the case speed is selected operator supports only 8 bit integer input images The grad operator is often used for edge enhancement with the magnitude related to the slope of the edge and the direction related to its orientation A faster approximation to the
400. rString Default 0 100 Type int Default 1 Values No 0 Yes 1 Type CorString Default y axis Type int Default 0 Values Auto 0 Manual 1 Type CorString Default 0 100 Type int Default 1 Values No 0 Yes 1 Type int Default 0 Values bitwise OR Fix position 1 X Type int Default 0 Y Type int Default 0 width Type int Default 500 height Type int Default 500 C Prototype CorOpRtn cor_graph CorObj In CorImage Graph CorString grTitle CorString name int type int cumulative int allowDuplicateNames int showDataPts int showLegend CorPoint legendPos CorString xaxis int axisXScale CorString xaxisRange int xShowTicks CorString yaxis int axisYScale CorString yaxisRange int yShowTicks int Placement int X int Y int width int height Headers include WITHOME h wDisplay h Libraries WITHOME lib wObj lib WITHOME lib wDisplay lib iEqualize Interactive histogram equalization Canis 4 uit ma ibug Description The iEqualize operator displays a grayscale image in a separate window and provides controls for selecting an appropriate contrast The effect of the ramp window changes can be seen immediately as the windowing arrows are moved When the desired equalized image is achieved click H the image you see is sent to the output f tie mepe Properties Main panel name Name of the window used to disp
401. racts run lengths from the image and outputs them in a vector in which run lengths from each contiguous set of pixels in the image are grouped together and given a distinct id number Images in the sequence are treated as if the last line of the previous image is adjacent to the first line of the current image and the first line of the next image is adjacent to the last line of the current image allowing blobs to extend across more than one image Blobs are output when they do not reach the bottom of the current image and so cannot extend into the next image in the sequence Blobs that reach the bottom row of the current image are usually saved by the operator to see if they extend into the next image in the sequence The operator can be forced to output these blobs using the reset parameter The operator is approximately equivalent to applying the getBlobs operator to an image formed by concatenating the images in the stream into a single image The difference is that this operator provides parameters to reset the operator s state and to limit the maximum height of the blobs produced Inputs BinaryImage The BinaryImage input must be an 8 or 16 bit unsigned image The images may be interpreted as either binary images such as those produced by the thresh operator or labeled images such as those produced by the kMeans operator All images received by an instance of the operator must be the same size and type Outputs Blobs The Blobs
402. rameter must be set to a value from zero to one where zero indicates no match between the pattern and target and one indicates a perfect match If no part of the target image matches the pattern sufficiently well empty output vectors are produced acceptThresh The acceptThresh parameter specifies an acceptance threshold above which a score value is always considered a valid match If the maximum number of matches specified by the maxMatches parameter are found with match scores above the acceptThresh value the search stops immediately This parameter is used to control early termination of the search for improved speed relThresh The relThresh parameter is used to set a relative rejection threshold for the matches The relative rejection threshold is determined by multiplying the parameter value by the match score of the best match found This value is then applied like the normal rejection threshold minContrast The minContrast parameter is used to specify the minimum contrast required in the target image to produce a match This parameter is usually set to a value between zero and one although values greater than one can be used if the target image has higher contrast than the model images The Candidates Sub Panel The patFind operator uses a two level seach strategy to speed up the matching process First a lower resolution search is performed over the entire ROI to find candidate matches The candidate matches are
403. rameter to zero will result in vertices only being added on the side that contains the maximum deviation from the line segment connecting the polyline endpoints The approach when the input is a polygon is similar with the initial approximation containing two segments The segments connect the starting vertex of the polygon to the most distant vertex and the most distance vertex to the original point The algorithm then proceeds as for polylines When a single side adjustment is specified for polygons right adjustment gives best results for polygons drawn in the counter clockwise direction and left adjustment gives the best result with polygons drawn in the clockwise direction Vertices in polygons created from perimeters using the plotFeatures operator are ordered clockwise so the left setting usually provides the best results Parameter Information max Deflection Type int Default 100 deflection Type Type int Default 0 Values mean 0 max 1 adjust Type int Default 0 Values both 0 left 1 right 2 niterations Type int Default 5 Demo iGraph Operators Geometry convexHull igr C Prototype CorOpRtn cor_reduceLines CorObj Points CorObj LinePoints int max_Deflection int deflection_Type int adjust int nIterations Headers include WITHOME h wFit h 111 Libraries WITHOME lib wObj lib WITHOME lib wFit lib Control Flow control operators for WiT igraphs buf Object bu
404. rcentage of the total image area falls Values default 0 char 1 uchar 2 between each value and the corresponding end of short 3 ushort 4 int 5 uint 6 the histogram The percentage is controlled through float 7 double 8 the fraction parameter Allowable range is 0 50 conversion Type int Default 0 Parameter Information Values round 0 floor 1 ceiling 2 fraction Type float C Prototype Default 5 CorOpRtn calcl Demo iGraph Ce ne CorObj Out CorString expression Operators Calculate Lut double userConstant int type int conversion C Prototype CorOpRtn cor_autoEqualize CorImage In Headers CorImage Out ee D yee include WITHOME h wCalcul h Headers Libraries include WITHOME h wCalcul h WITHOME lib wObj lib WITHOME lib wCalcul lib Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib 14 calc Perform arithmetic and logic operations on two input objects Description The calc operator performs ALU type operations on two input objects The input objects may be any combination of primitive objects or vectors of primitives The output is cast to integer or double following the same casting rules as in the C programming language The operation to be performed is specified by a C style mathematical expression entered using the expression parameter The userConstant field allows the user to specify a constant of type double for use in the specified expressio
405. ream into a single image The difference is that this operator provides parameters to reset the operator s state and to limit the maximum height of the blobs produced and perimeters cannot be extracted by this operator Inputs BinaryImage The BinaryImage input must be an 8 or 16 bit unsigned image The images may be interpreted as either binary images such as those produced by the thresh operator or labeled images such as those produced by the kMeans operator All images received by an instance of the operator must be the same size and type ROI The ROI input must be single rectangle graphic or a vector of rectangle graphics The y coordinates of the rectangular ROI are specified relative the the frame of reference of the stream of images rather than that of the individual input images ROI The ROI input must be single rectangle graphic or a vector of rectangle graphics The y coordinates of the rectangular ROI are specified relative the the frame of reference of the stream of images rather than that of the individual input images Outputs BlobFeatures The BlobFeatures output is a Blob feature vector Each element of the vector contains information related to one contiguous region in the input image This information includes such things as the number of pixels in the region and its centroid The Blob feature structure comprises the following fields 272 Field Description Name id Unique blob identifier npixels Number o
406. res operator The Feature id is the same as blob id from corresponding blobs in the original image A typical application selects a Feature or Features from the Feature vector using the selectElem or partition operator and applies the resulting Features to the selectBlobs operator as an input parameter to obtain the corresponding blobs Parameter Information select Type CorObj Demo iGraph Operators Blobs BlobStats C Prototype CorOpRtn cor_selectBlobs CorBlobVector Blobs CorBlobVector Selected CorBlobVector NotSelected CorObj select Headers include WITHOME h wBlob h Libraries WITHOME lib wObj lib WITHOME lib wBlob lib Calculate The Calculate library provides a range of functions based on arithmetic and logical calculations The library includes operators that perform arithmetic and logical operations on images vectors and numeric values operators that equalize images operators that average images and operators that produce random or deterministic numbers or number sequences add Add two objects Description Add two objects The objects can be simple scalar values such as integers or floating point numbers or vectors or images of such numbers The type of the output is promoted to the larger of the two input types if they are different For example if one of the inputs is a float and the other one an 8 bit integer the result will be a float 10 C Prototype CorOpRtn cor_add
407. resolution The parameters on the Candidates subpanel are used to control the low resolution search and candidate selection scale The scale parameter is used to specify resolution reduction factor for the initial search for candidates It can be set to powers of two from two to 32 max Candidates The maxCandidates parameter specifies the maximum number of candidates that will be passed to the second phase of the search This should usually be greater than the number of matches required in the second phase candidateThresh The candidateThresh parameter is used to set the rejection threshold for candidate matches candidateRelThresh The candidateRelThresh parameter is used to set a relative rejection threshold for candidate matches The relative rejection threshold is determined by multiplying the parameter value by the match score of the best candidate match found This value is then applied like the normal rejection threshold The relative threshold is used to restrict the candidates passed on to the second to those near the best applyPatterns The applyPatterns parameter is only meaningful when more than one pattern model is applied When the parameter is set to All all patterns in the input pattern vector are compared to the target image in both phases of the search When the parameter is set to Sample the pattern vector is sampled as specified by the patSampling parameter for the candidate search The second phase of the
408. ring 0 CharVector 1 C Prototype CorOpRtn cor_readSerial CorObj Data int port int maxChars int output Headers include WITHOME h wSerial h Libraries WITHOME lib wObj lib WITHOME lib wSerial lib writeSerial Write to a serial port 4 ne Description The writeSerial operator writes the input string or characters to a previously opened serial port The port parameter specifies the serial port written to The port parameter value 0 corresponds to the COMI port parameter value 1 corresponds to COM2 and so on The port must have previously been opened using the openSerial operator Attempting to write to a port that has not been opened results in an error The input may be a string or a vector of characters WiT types char byte and ubyte will all be accepted The operator will write characters from the input to the port until all the characters have been written or until a write timeout is reached Timeouts are set when the port is opened by the openSerial operator The operator outputs the number of characters successfully written to the port Parameter Information port Type int Default 1 C Prototype CorOpRtn cor_writeSerial CorObj Data int NumWritten int port Headers include WITHOME h wSerial h Libraries WITHOME lib wObj lib WITHOME lib wSerial lib Video Acquisition acqColorkey Set color key index for live video passthrough kg omen 1m
409. rmation derived from them are referenced to the beginning of the first image in the sequence of images Parameters connected The connectivity of the output blobs is specified by the connected parameter When this parameter is set to 4 pixels that share a common edge are considered to be connected so each pixel has four neighbours When it is set to 8 pixels that touch corner to corner are also considered connected so each pixel has eight neighbours type The type parameter determines how the input image will be interpreted When it is set to binary all non zero pixels are considered to be equivalent and the run lengths will represent runs of non zero pixels When it is set to labeled pixels with the same value are considered equivalent and run lengths represent runs of equal valued non zero pixels maxBlobHeight The maxBlobHeight parameter sets an upper limit on the height of blobs produced by the operator When the information derived using the webGetBlobFeatsRoi operator is used with images produced by the webWaterfall operator the maxBlobHeight value should be less than the height parameter value of the webWaterfall operator by at least the height of the images in the input image stream to ensure that the blobs are contained within a single waterfall image reset The reset parameter can be used to release all active blobs and reset the running line count to zero at the end of the current image When a reset function
410. rom them are referenced to the beginning of the first image in the sequence of images Parameters connected The connectivity of the output blobs is specified by the connected parameter When this parameter is set to 4 pixels that share a common edge are considered to be connected so each pixel has four neighbours When it is set to 8 pixels that touch corner to corner are also considered connected so each pixel has eight neighbours type The type parameter determines how the input image will be interpreted When it is set to binary all non zero pixels are considered to be equivalent and the run lengths will represent runs of non zero pixels When it is set to labeled pixels with the same value are considered equivalent and run lengths represent runs of equal valued non zero pixels maxBlobHeight The maxBlobHeight parameter sets an upper limit on the height of blobs produced by the operator When the information derived using the webGetBlobs operator is used with images produced by the webWaterfall operator the maxBlobHeight value should be less than the height parameter value of the webWaterfall operator by at least the height of the images in the input image stream to ensure that the blobs are contained within a single waterfall image reset The reset parameter can be used to release all active blobs and reset the running line count to zero at the end of the current image When a reset function is required this parameter
411. rosses the slope threshold The edge location is set to the point at half value between the two threshold crossings When the edgeLocation parameter is set to MaxSlope the edge is located at the point of maximum slope and the start and end of the edge are the points on the path on either side of this point where the slope crosses zero or changes direction from increasing to decreasing or decreasing to increasing whichever is nearer In both cases the slope value is set to the maximum slope over the edge and the level value is the absolute value of thelevel change between the start and edge of the end For peak edges the edge location is the zero crossing of the path slope The edge direction is parallel to the path at the crossing point in the direction of the path for rising step or peak edges in the opposite direction for falling step edges The slope value assigned to peak edges is the slope of the path slope approximating the second derivative in the path direction where the path slope crosses zero The level is the image value on the path at the zero crossing point Two threshold values may be used to extract edges based on the slope and the level of the edge The criterion used for each of the thresholds depends on whether the edges are step edges or peak edges For step edges the slope criterion is the slope of the edge and the level is the step change in value over the edge For peak edges the slope criterion is the sharpnes
412. rs include WITHOME h wWeb h Libraries WITHOME lib wObj lib WITHOME lib wWeb lib webLabelBlob Rasterize binarize outline blobs in a waterfall image sahi en et Sed Description The webLabelBlob operator produces images from Blob vectors The image may be produced in a frame of reference that is shifted in y by a specified amount The image produced may be a binary image or a labeled image of the blobs in the input vector or a binary outline of the blobs in the the input vector The original image that was fed to getBlobs is required at the top input to create the correct output image size Inputs BlobSource The BlobSource input must be an image This image defines size of the output 281 image and is usually the image from which the blobs were extracted Blobs The Blobs input must be a blob vector For this operator the blob vector is usually produced by the webGetBlobs operator Outputs LabeledImage The output image is an unsigned eight bit image the same size as the input image The content of the image is determined by the setting of the labelT ype parameter Parameters offset The value in the offset parameter is subtracted from the row number of the run lengths in the Blob vector before the output images are produced so the images appear shifted up by that amount in the output image labelType The labelType parameter determines how the output image is formed If it is set to greyscale the output is
413. s C Prototype CorOpRtn cor_kMeans CorImage In CorImage Out CorVector Mean CorVector Variance CorVector Number int initMeans CorIntVector meanValues int numClusters int maxIterations Headers include WITHOME h wSegment h Libraries WITHOME lib wObj lib WITHOME lib wSegment lib localAdaptThresh Threshold an image relative to local mean pixel value Description The localAdaptThresh operator applies thresholding to a grayscale or color image using thresholds based on local pixel values in the input image For grayscale images the thresholds are applied directly to the values of image pixels For color images the thresholds are applied to a luminence or intensity value for each pixel the Y channel for Yuv images the V channel for HSV images and a luminence value derived from all three channels equivalent to the Y value for RGB images The operator function is similar to that of the thresh operator except that values corresponding to the lo and hi parameters of the thresh operator are calculated for each pixel in the image based on pixel values in the its neighbourhood Low and high threshold values are calculated for each pixel in the image Each threshold value is set to the local mean pixel value plus an offset The local mean is calculated over a rectangular region around the pixel whose size is specified by the width and height parameters The method of calculating the offsets
414. s goto Send token to specified label Description The goto operator allows tokens to be sent to a distant location without a long link on the igraph This is particularly useful when an igraph is very cluttered The instance name of label is used to match the destination specified in the label parameter of goto A wild card may be used to send an object to multiple label operators Parameter Information label Type CorString Default label Demo iGraph Operators clearObjs if Conditional branch Description The if operator provides conditional branching ability in an igraph It has one boolean parameter which like all other parameters can be supplied from other igraph operators If the condition is true the input token is passed to the top branch labelled with TrueCond Otherwise the input token is passed to the bottom branch labelled with FalseCond See also if2 if3 if4 Parameter Information cond Type int Default 0 Values False 0 True 1 Demo iGraph Operators clearObjs if2 Conditional branch for 2 inputs Description Similar to the if operator but controlling 2 objects with the same condition See also if if3 if4 Parameter Information cond Type int Default 0 Values False 0 True 1 Demo iGraph Operators clearObjs if3 Conditional branch for 3 inputs J eae Description Similar to the if operator but controlling 3 objects with the same condition
415. s The user may choose the mode by using the optimization parameter If the parameter value is speed then execution goes way faster otherwise operator delivers more precise results The operator accepts integer and float type input images and color images in the case optimisation is speed An unsigned integer input image produces a signed integer output image with the same number of bits as the input image In all other cases the output image is the same type as the input image 86 Parameter Information width Type int Default 3 height Type int Default 3 optimization Type int Default 1 Values preciseness 0 speed 1 Demo iGraph Operators Filter Entropy Operators Filter V ariance C Prototype CorOpRtn cor_hipass2d CorImage In CorImage Out int width int height int optimization CorOpRtn cor_hipass2d_consume CorImage In int width int height int optimization Headers include WITHOME h wFilter h Libraries WITHOME lib wObj lib WITHOME lib wFilter lib laplace Laplacian 2nd order edge enhancement Description The laplace operator enhances changes in local contrast by applying a filter that approximates a laplacian second derivative operation to the image The filter subtracts an average of neighbouring pixels from each pixel in the image emphasizing points lines and edges and suppressing values in regions of constant gradient Three types of laplacian filter c
416. s mean standard deviation minimum and maximum of the image pixel values within the image ROI or of the vector element values are computed and produced on the top to bottom output ports respectively If the input is an image and he useMask parameter is set to Yes the statistics are calculated using pixels in the intersection of the image ROI and a mask specified by the Mask parameter The mask can be specified by a Geometry or Graphic object or by a binary image The mask may also be specified by a vector of Geometry or Graphic objects binary images or Blobs In this case pixels may be counted more than once if the elements overlap in the image Parameter Information useMask Type int Default 0 Values No 0 Yes 1 Mask Type CorObj Default OBJ_B T CorGeom 4 CorFpoint 2 0 0 99 99 Demo iGraph Operators Statistics Stats Operators Blobs BlobStats C Prototype CorOpRtn cor_stats_1 CorObj In float Area float Mean float Var float Min float Max int useMask CorObj Mask Headers include WITHOME h wStats h Libraries WITHOME lib wObj lib WITHOME lib wStats lib File IO The File IO library supports writing to and reading from files The library includes operators that read and write WiT objects operators that read and write a number of image formats and operators to support related functions CPUTime Get the current CPU time Description Get the CPU time typicall
417. s a vector of vectors of graphics corresponding to each suspicious region found in the original image This is sometimes useful for debugging Test The Test output is an idealized image of the found ECC200 barcode Parameters cellSize The cellSize parameter determines the size of each cell comprising the 2D bar code given in pixels This size must be bigger than the distance between two cells or dots in a barcode but smaller than the silent zone around the perimeter of the barcode Parameter Information cellSize Type int Default 10 Demo iGraph Operators Matrix ECC200 igr C Prototype CorOpRtn cor_getECC200 CorImage Target CorStringVector Decoded CorImageVector Test CorGraphicVector Info int cellSize Headers include WITHOME h wMatrix h Libraries WITHOME lib wObj lib WITHOME lib wMatrix lib getECC200_1 Find and decode ECC 200 2D barcodes EN Description The getECC200_1 operator recognizes ECC200 barcode See getECC200 help for details Differ from getECC200 output of this operator is a vector of vectors of bytes rather then the vector of CorStrings Parameter Information cellSize Type int Default 10 moduleSize Type int Default 10 Demo iGraph Operators Matrix ECC200 igr C Prototype CorOpRtn cor_getECC200_1 CorImage Target CorVector DecodedBin CorGraphicVector Info int cellSize int moduleSize Headers include WITHOME h wMatrix h Li
418. s controlled by the result parameter If result is binary the output is an 8 bit unsigned image with on pixels set to 255 and and off pixels set to 0 If result is semi threshold the output image has the same type as the input image On pixels maintain their original values while off pixels are set to the value specified by the semiFillValue parameter For scalar input images this parameter should be set to a scalar value If the value is outside the range of the input image type the value will be clipped For color images the semiFillValueparameter can be set to a the same type as the input image or to a scalar value If it is 195 set to a scalar the value is clipped to the unsigned byte range and applied to all channels If levels is double and invert is No pixels in the range lo hi inclusive are turned on otherwise off Setting the invert parameter to Yes causes the on and off ranges to be flipped If levels is single pixels equal to or above lo are off If levels is double pixels in lo hi inclusive are off Parameter Information lo Type float Default 0 hi Type float Default 255 invert Type int Default 0 Values no 0 yes 1 levels Type int Default 1 Values double 0 single 1 result Type int Default 1 Values semi threshold 0 binary 1 semiFillColor Type CorObj Default OBJ_B T int 0 Demo iGraph Operators Segmentation Thresh Operators Segmentation
419. s forming the corner fully visible in the image top left top right bottom left bottom right Notice that because natural image coordinate system has the X axis down and origin at the upper left corner the image will appear to be flipped depending on the origin setting The scale factor is kept in such a way that measurements in the original image and destination image are as close as possible This way the corrected image and data points can be displayed as an image similar in size to the original one The proportion output is the ratio between physical units corresponding to the cellSize paramter and pixels If measurements in physical coordinates are needed they can be computed by multiplying positions and lengths in the corrected image by proportion The proportion value can also be used for the calibrateScale operator so that the correct pixel to physical coordinate scaling can be applied automatically when restoreCoordinates is used Parameter Information cellSize Type float Default 1 interpolation Type int Default 0 Values perspective 0 bilinear 1 origin Type int Default 0 Values none 0 bottomLeft 1 topLeft 2 topRight 3 bottomRight 4 Demo iGraph Operators Calibration calibrate igr C Prototype CorOpRtn cor_calibrateCamera CorImage chessBoard CorCalibrateInfo calibrateInfo float proportion float cellSize int interpolation int origin Headers include WITHOME
420. s from the image The range of scales is specified by the minScale and maxScale parameters As with the rotation range the scale range is limited before matching so characters for which the best match is outside the scale range may be considered a match if the scale limited match is sufficiently good minScale and maxScale These parameters specify the minimum and maximum value of the scale correction factor applied to information extracted from the image when the scaleInvariant parameter is set to No threshold The threshold parameter allows the specification of a minimum acceptable match quality This parameter is set to a value from 0 to 1 where a setting of zero means that every blob will be assigned some matching character from the font and a setting of 1 requires a perfect match The treatment of blobs for which sufficiently good match is not found is controlled by the ignoreNoMatch and noMatchLabel parameters ignoreNoMatch When this parameter is set to Yes image blobs that do not match a character in the font model that is those for which the best match quality falls below the threshold value are ignored When the parameter is set to No such blobs are assigned the label specified in the noMatchLabel parameter noMatchLabel The noMatchLabel String parameter specifies the label that is assigned to blobs that do not match a character in the font model when the ignoreNoMatch parameter is set to No selection The selec
421. s in the image threshold The threshold parameter sets the threshold used when the caleThresh parameter is set to No threshFactor When the calcThresh parameter is set to Yes the threshold is based on the image mean and minimum pixel values and is controlled by the threshFactor parameter The threshFactor parameter 234 is usually set to a value between 0 and 1 A value of 1 will give the image mean as the threshold value Lower values produce thresholds nearer the foreground value and usually result in thinning of the blobs in the binarized image Parameter Information background Type int Default 0 Values light 0 dark 1 calcThresh Type int Default 1 Values No 0 Yes 1 threshold Type float Default 0 threshFactor Type float Default 0 7 C Prototype CorOpRtn cor_ocrBin CorImage In CorImage Out float Threshold int background int calcThresh float threshold float threshFactor Headers include WITHOME h wOcr h Libraries WITHOME lib wObj lib WITHOME lib wOcr lib ocrEdit Edit OCR parameters Description Parameter Information Placement Type int Default 0 Values bitwise OR Fix position 1 X Type int Default 0 Y Type int Default 0 C Prototype CorOpRtn cor_ocrEdit CorObj In CorOcrinfo Out int Placement int X int Y Headers include WITHOME h wOcr h Libraries WITHOME lib wObj lib WITHOME lib wOcr lib ocrGraylnit
422. s may be ridges a line of pixels higher than the surrounding pixels with the resulting enhanced line positive or troughs a line of pixels lower than the surrounding pixels with the resulting enhanced line negative The magnitude of the enhanced line depends on the orientation of the line relative to the specified orientation and the difference between the pixels on the line and the adjacent pixels The line orientation is specified as one of E W NE SW N S or NW SE with the direction parameter A second derivative filter is applied to the image to detect changes in the image gradient in the direction perpendicular to the orientation The magnitude of the resulting image is greatest for pixels which are part of lines with the specified orientation and falls to zero for pixels that are not part of a line or are 87 part of a line perpendicular to the orientation The operator responds most strongly to ridge or trough type lines but also enhances step edges The operator accepts integer and float type input images An unsigned integer input image produces a signed integer output image with the same number of bits as the input image In all other cases the output image is the same type as the input image The output is scaled so that for integer images a full scale difference between a line in the specified orientation and its surrounding pixels will produce a line in the output image with the maximum image value Parameter Inform
423. s may be written in binary or text format as specified by the format parameter The binary format is recommended for images and other large data objects because the resulting files are smaller The text format files are useful when the objects are to be edited outside of WiT or when they must be read by other applications Parameter Information filename Type CorFile Default outfile format Type int 216 Default 0 Values binary 0 text 1 none 2 display 3 C Prototype CorOpRtn cor_writeObj CorObj In CorFile filename int format Headers include WITHOME h wSystem h Libraries WITHOME lib wObj lib WITHOME lib wSystem lib Transforms The Transforms library includes operators implementing the fast Fourier fast Hadamard and discrete cosine transforms det Discrete cosine transform of an image Description The det operator computes a forward or inverse discrete cosine transform of the input image The transform is implemented using an FFT as an intermediate step The input image may be an integer or float image it may not be complex or color RGB The operator s output is a float image with the size of the output image equal to that of the input image regardless of the input image ROI size rounded up to the nearest power of two in each dimension Only pixels in the input image ROI are used in the DCT calculation A forward or inverse cosine is specified by the dir parameter
424. s of the peak and the level criterion is the actual value at the peak The threshold values are set using the slopeThresh and _levelThresh parameters Each threshold value can be either relative or absolute as specified by the slopeType and levelType parameters Relative thresholds are specified as a percentage of the maximum value of the corresponding quantity found on the path while absolute thresholds are based on the actual value of the quantity The operator also produces vectors containing the sampled path values and the smooothed slope values along the path These vectors can be displayed using the graph operator They can be helpful when determining settings for the thresholds and other parameters of the operator Parameter Information edgeType Type int Default 2 Values rising 0 falling 1 rising or falling 2 peak 3 edgeLocation Type int Default 0 Values HalfLevel 0 MaxSlope 1 75 slopeType Type int Default 0 Values relative 0 absolute 1 slopeThresh Type float Default 20 levelType Type int Default 0 Values relative 0 absolute 1 levelThresh Type float Default 50 sampleSpacing Type float Default 1 smoothing Type int Default 1 Demo iGraph Operators Edges EdgesOnCircle C Prototype CorOpRtn cor_getXEdges CorImage In CorObj Path CorEdgeVector Edges CorFloatVector PathValues CorFloatVector SlopeValues int edgeType int edgeLocation int slopeType fl
425. sRemoved output The if true branch is connected to the skeleton operator input and the if conditional parameter port is connected to the pixelsRemoved output port Skeletons will appear out the false branch when pixelsRemoved reaches zero By setting a probe on the output of the skeleton operator one can watch the image gradually collapse to its skeletal form Parameter Information numPasses Type int Default 1 Demo iGraph Operators Morphology skeleton igr C Prototype CorOpRtn cor_skeleton CorImage In CorImage Out int PixelsRemoved int numPasses Headers include WITHOME h wMorpho h 152 Libraries WITHOME lib wObj lib WITHOME lib wMorpho lib thicken Binary thickening filter i af ft z Description The thicken operator performs a thickening type binary pattern matching operation on an image If the pixels in the neighbourhood of a given pixel match the pattern of pixels specified in the kernel then the corresponding pixel in the output image is set to the image maximum otherwise it is unchanged The kernel parameter allows the user to specify the pattern Kernel elements set to 0 match 0 value pixels in the input image kernel elements set to 1 or any positive integer match non zero pixels in the input image and kernel elements set to 1 or any negative integer match all pixel values in the input image The input and output of the thicken operator are both 8 bit unsigned i
426. sed to specify whether the edge point search uses interpolation to find a sub pixel location for the edge points before fitting to the edge line The subpixel parameter Parameter Information measure Type int Default 0 Values PO P1 0 PO E1 perp to EO 1 PO E1 shortest 2 MidPoint 3 boxWidthO Type int Default 10 boxHeightO Type int Default 10 numPathsO Type int Default 3 boxWidth Type int Default 20 boxHeightI1 Type int Default 20 numPaths Type int Default 3 edgeTypeO Type int Default 2 Values rising 0 falling 1 unknown 2 slopeThreshO Type float Default 1 levelThreshO Type float Default 1 edgeTypel Type int Default 2 Values rising 0 falling 1 unknown 2 slopeThresh1 Type float Default 1 levelThreshI Type float Default 1 subpixel Type int Default 1 Values No 0 Yes 1 C Prototype CorOpRtn cor_caliper CorImage i0 CorObj RefLine CorEdgeVector Edges float Dist CorGraphicVector Points int measure int boxWidth0o int boxHeight0 int numPaths0O int boxWidthl int boxHeightl int numPaths1 int edgeTyped float slopeThreshO float levelThreshO int edgeTypel float slopeThreshl float levelThreshli int subpixel Headers include WITHOME h wMeas h Libraries WITHOME lib wObj lib WITHOME lib wMeas lib distAlIPtToPt Find the distance between all pairs of points Description The distAlIPtToPt op
427. sed to specify which element is extracted with the first element indexed by zero The extracted element is produced on the upper output port the remainder of the vector or list on the lower output port Parameter Information method Type int Default 0 Values first 0 last 1 byElement 2 element Type int Default 0 C Prototype CorOpRtn cor_getElem CorObj Vector CorObj newElement CorObj newVector int method int element 162 Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib getField Extract a field from input object al he k ri Description The getField operator allows the user to extract a field from a complex object or vector of complex objects Object types from which fields may be extracted include the built in types Point WitImage Vector and Llist extended types such as Fpoint Graphic and Feature and user defined types The field to be extracted is specified by name in the field parameter using a syntax similar to the C language Fields in the WiT built in types that may be of use include x and y in Point W image width H image height bpp bytes per pixel and type in image and data size ncols and type in vectors The data field refers to the pointer to memory containing the elements and is most often used with an index value to specify one element in the vector The field name data n specifies element n of the vector An
428. seeeneeneesees COUNLP IX ii EE E E ETOR 199 getRSS14 differente sessed sees E E EE S a 199 getUPCa BetSIOPCMAXIMA 0 ceeeecesccseeseeseeseeseeeeneeteeneenees 200 PAUP CO asiti e E E E ER histCentroid seinra E R 201 gradeECC200 VAA L01 EERE ET EEEE 201 warpMatrix histogram ariii reiri isr arta iK E 202 SMAMtOCR sicciscecccenccatasstesedssecorssaeossdertsenessbaseteseiers PTO OCH ON sai evita Noein eon tee OCT BIN EENE cass EEEE E EEE EST smoothHist ocrEdit ocrGraylnit ocrGrayRun ocrLocalBin OCTMAKE Font nenons iii na 239 ocrMakeSemiFont s es 240 ocrPerimInit fileStat s aen an ne a RA A ie 206 OCT POTIMRUM e neri e E RR 242 getClipboard non ne ice a ees 207 ocrPlot Font serosa si 245 getTime ocrRoiBin WAKA ROA EE E E ET OCTRUM e na SE A EEA e Na NSE playSound oooseeseeeeeeeeeeeesereeerersrsrerrereser ocrSelectFont ssseeseeeseseeeeeeesererrrrserererrereseses 248 p intemannass SmartSearch esisiini rieri reneis 248 putClipboard patFkind leven awe REEE ETA A 248 TAT CX eeeseeseee DGLEING2Z DI ores eke een A Ea ks 252 readImage POLNIET ioc wels renen sees ci nas Matas eerie atlas eth aSa 255 readOJj POEINUZ DL ver siete vncass seuss ta deseia sed edunde seebedenes cheeses 257 ringBell PAtInterPOlAte neninnn ennn naa 259 splitName patoca Uinise 259 stopSound POLROL 2 cst a a a E dtessteametsh 262 stopWatch PATRON eain esorta ERa cesses thavi
429. slope zero crossing or local minimum on either side of the slope maximum The direction field contains the direction of the line The operator also produces vectors containing the sampled path values and the slope values along the path These vectors can be displayed using the graph operator They can be helpful when determining settings for the thresholds and other parameters of the operator The edgeType parameter determines if the operator looks for rising edge falling edges or both The subpixel parameter can be used to specify the use of interpolation along the sample line to obtain a sub pixel estimate of the edge location 77 Values are extracted from the input image at one pixel intervals along the specified line The lineWidth parameter determines how many pixels on either side of the line are used to calculate these values The pixel values are weighted by a linear decreasing function of their distance from the line The lineWidth parameter value is the half value width of the linear weighting function A lineWidth value of 0 gives nearest neighbour sampling and a value of J gives linear interpolation using the pixels nearest the line The smoothing parameter allows a moving average filter to be applied to the sampled path values The value of the parameter specifies the width of the filter in pixels Parameter Information edgeType Type int Default 0 Values rising 0 falling 1 rising or falling 2 slopeThresh
430. sssrrisirssrresssss 32 andBlobs 1l m yO rigita nann aeo ARE EE 32 getBlobFeatures a TESTOTE COOTAINALES eerren 33 getFeatures 4 restorelmage BetPeCviMeleL esceeeee 5 Color getWarpedFeatures sin bayer getWarpedPerimeter 6 conVertCOlor asrda E 35 labelBlobs vescecscesescscssees A convertColorFormat mergeFeatures STR BOLCNANNEL ainnise a a r eaadas minSizeFeatures 9 me rgeChannel aceon na aE SCLCCIBIODS 3 222 tok sh gunscass ie tothe 9 Cale late ariete eisiea eras air chdntasveducased 10 LA A EEEE EAE ESIET TE 10 CoMmpleX iyare aiar h 38 GAANOIS Cs n E ees aE 10 complexMerge E AAE E entice eee ee 11 COMPLEXSDLit v sessessesereeseeeresseeeteeneneeeenenteneeeneens 39 SPCCU UM e e Se E a E R EE E 39 Convert lt Vypeitsccs ceiecesistesitics R R iS 40 collectAverage sf RSE eee A E A ne RAE Ten 42 comparelmages ascseccccsescsessssevsesesseseseseseeseseeseaees 16 CVC AS Opr R EEO ERE EOE E ERE 43 complxMult JormatStri ng zonenn anna a a etna 43 conditional v ecccccccceseccesscessccesscessscesscessecessecesscees plotkdges see e E E cust E EE R 44 diVar a a E A a ERA plotFeatures usssssssssssssisssissssssssssessesrsrrssssesssssns 44 equalize false DIOLL ANOS onana EE E A R R 46 fastAluOp SCONS IVINS ai icicles abs tes Helen tended S eaaa 47 Fast UNAYO D A raa nial eee a splitString NIStEQUGLIZE 8 2 ana nE a EER 22 STUN EA E E E sev eas oa aces todas eeaeiees TAA
431. subpixel mode of the patInit operator resolution The resolution parameter specifies the resolution of the subpixel refinement of the match locations The resolution can be specified in powers of two from 2 half pixel to 64 1 64 pixel This parameter is only used when the subPixel parameter is set to Yes The Candidates Sub Panel The patFind operator uses a two level seach strategy to speed up the matching process First a lower resolution search is performed over the entire ROI to find candidate matches The candidate matches are then refined at full and possibly subpixel resolution The parameters on the Candidates subpanel 253 254 are used to control the low resolution search and candidate selection scale The scale parameter is used to specify resolution reduction factor for the initial search for candidates It can be set to powers of two from two to 32 max Candidates The maxCandidates parameter specifies the maximum number of candidates that will be passed to the second phase of the search This should usually be greater than the number of matches required in the second phase candidateThresh The candidateThresh parameter is used to set the rejection threshold for candidate matches candidateRelThresh The candidateRelThresh parameter is used to set a relative rejection threshold for candidate matches The relative rejection threshold is determined by multiplying the parameter value by the match s
432. t given by the userConstant parameter and by the length of the input vector The symbol K is used to represent the user constant and the symbol W is used to represent the input vector length in the expression The expression syntax is described in the operator help for the calc operator The calculations are performed using floating point arithmetic with the final index value rounded to the nearest integer 5 rounded up The floor ceil and round functions can be used in the expression to control truncation of the floating point values if necessary Calculated index values less than zero 169 are set to zero and calculated index values greater than the maximum valid input index are set to the maximum valid input index Note that the expression is a function of the output indexes and the function resolves to an input index value It is often more natural to think of the output index as a function of the input index but the method used here allows the input to output mapping to be more general allowing many to one mappings and subsets of the input and also avoids problems with unspecified output vector elements The length of the output vector is specified by a second expression given in the lengthExpr parameter This expression may use the user constant K and input vector length W The index value A has no effect on the value of this expression otherwise the syntax is the same as for the expression parameter An integer
433. t int aluOp int sizeOp CorHex overlayMask CorPoint offset CorString expression int outputType Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOME lib wCalcul lib applyLut Apply a lookup table on an image thee sof F Description The applyLut operator applies a lookup table LUT received on the bottom input port to an integer image received at the top input port The result is to map each pixel value z in the image Z to anew pixel value Z specified by a precalculated mapping function f to generate a new image J Z fl The input image must be one of the integer image types unsigned byte signed byte unsigned short or signed short The input LUT may be a signed byte unsigned byte signed short unsigned short or float vector To maintain compatability with earlier versions of WiT the input LUT may also be a 32 bit signed integer vector The LUT must include sufficient elements to cover the range of the input image 256 elements for 8 bit images and 65 536 elements for 16 bit images A LUT in the correct format can be produced by the makeLut operator Except in the case where the input LUT is a signed int vector the output image is the same type as the input LUT For unsigned input images the input image pixel values are used directly as indexes into the LUT vector The indexed value from the LUT is assigned to the corresponding pixel in the output image
434. t 255 Values 0 255 fillG Type int Default 255 Values 0 255 fillB Type int 181 Default 0 Values 0 255 C Prototype CorOpRtn cor_rasterize CorImage Image CorObj Graphics CorImage Rasterized int overlay int rangeType float min float max int outType int useColor int outline int penval int penR int penG int penB int fill int pixval int fillR int fillG int fi11B Headers include WITHOME h wPixel h Libraries WITHOME lib wObj lib WITHOME lib wPixel lib sampleLine Extract values along a line in the image Description The sampleLine operator extracts pixel values along a line in an image The inputs to the operator are an integer or float image at the upper input port and line type graphic object at the Line input port Values are extracted from the input image at one pixel intervals along the specified line The lineWidth parameter determines how many pixels 182 on either side of the line are used to calculate these values The pixel values are weighted by a linear decreasing function of their distance from the line The lineWidth parameter value is the half value width of the linear weighting function A lineWidth value of 0 gives nearest neighbour sampling and a value of 7 gives linear interpolation using the pixels nearest the line The Points output is a vector of real points corresponding to the sample locations on the line The Values outp
435. t Default 0 Values None 0 2x2 1 3x3 2 5x5 3 7x7 4 minsize Type int Default 8 C Prototype CorOpRtn cor_burtStep CorImage In CorImage Gaussian CorImage Laplacian int Final int filter int minsize Headers include WITHOME h wPyramid h Libraries WITHOME lib wObj lib WITHOME lib wPyramid lib decimate Decimate an image Description The decimate operator takes an input image and reduces its size by half The new image consists of every fourth pixel taken from the input image The pixels used to compose the new image are taken from the decimation lattice defined by the useRows and useColumns parameters These two parameters can be set to either even or odd indicating that pixels should be taken from the even or odd rows columns of the input image No filtering is performed by decimate Parameter Information useRows Type int Default 0 Values Even 0 Odd 1 useColumns Type int Default 0 Values Even 0 Odd 1 C Prototype CorOpRtn cor_decimate CorImage In CorImage Out int useRows int useColumns Headers include WITHOME h wPyramid h Libraries WITHOME lib wObj lib WITHOME lib wPyramid lib expand Expand and filter an image 185 i i Description The expand operator increases the size of an input image by inserting a new row column of zero valued pixels between each row column of the input image The resulting i
436. t fontSize and fontStyle parameters Parameter Information graphicType Type int Default 0 Values Point 0 Line 1 Polyline 2 Polygon 3 Rectangle 4 Circle 5 Text 6 Arc 7 points Type Corlmage Default 0 0 outline Type int Default 1 Values off 0 on 1 penR Type int Default 255 Values 0 255 penG Type int Default 0 Values 0 255 penB Type int Default 0 157 Values 0 255 fill Type int Default 0 Values off 0 on 1 fillR Type int Default 0 Values 0 255 fillG Type int Default 0 Values 0 255 fillB Type int Default 0 Values 0 255 text Type CorString Default font Type int Default 0 Values helvetica 0 times 1 courier 2 JontSize Type int Default 12 fontStyle Type int Default 0 Values normal 0 bold 1 italic 2 bold italic 3 C Prototype CorOpRtn cor_crGraphic CorObj Graphic int graphicType CorIntImage points int outline int penkR int penG int penB int fill int fillR int fillG int fil1B CorString text int font int fontSize int fontStyle Headers include WITHOME h wObjman h 158 Libraries WITHOME lib wObj lib WITHOME lib wObjman lib crimage Create a image Description The crImage operator generates a new image every time an object is received at its input Image values can be set to a constant or to a ramp in the x or y direction The
437. t height CorIntImage kernel int optimization Headers include WITHOME h wMorpho h Libraries WITHOME lib wObj lib WITHOME lib wMorpho lib erode Morphological erosion Description The erode operator performs binary or gray scale morphological erosion on images Binary erosion removes pixels from the edges of objects in the image Gray scale erosion sets each pixel to the minimum pixel value in its neighbourhood and is sometimes refered to as a minimum filter This operator could be executed in two modes The user may choose the mode by using the optimization parameter If the parameter value is speed then execution goes way faster otherwise operator delivers more precise results Binary or gray scale erosion is selected using the type parameter The input for binary erosion must be an 8 bit unsigned image The result is a binary 8 bit unsigned image For gray scale erosion the input may be any integer or float gray scale image or color in the case optimisation is speed The output image is a gray scale image of the same type as the input image The kernelType parameter is used to select the type of structuring element or neighbourhood specifying kernel used The structuring element may be specified as a rectangular region with the size of the region is given by the width and height parameters or it may be specified as an arbitrary region in which case the neighbourhood is specified by the non zero el
438. t image is displayed with the graphic objects overlayed in color The source of the color used for the graphic objects is specified by the useColor parameter If it is set to original the color of each graphic object is taken from the penColor and fillColor fields in the object itself If useColor is set to forced a graphical object is drawn using the color whose red green and blue components are defined by the forcedR forcedG and forcedB parameters Rectangles 63 circles and polygons are not filled in the forced mode The output of overlayData is a color image as displayed by the operator Most of the operator properites of overlayData are the same as the display operator Parameter Information name Type CorString Default overlay background Type int Default 0 Values No 0 Yes 1 useColor Type int Default 0 Values original 0 forced 1 forcedR Type int Default 255 Values 0 255 forcedG Type int Default 255 Values 0 255 orce e int dB Type i Default 0 Values 0 255 showLabels Type int Default 0 Values Labels 0 None 1 Serial 2 Labels amp Serial 3 Placement Type int Default 0 Values bitwise OR Fix position 1 Max size 2 Min size 4 X Type int Default 0 Y Type int Default 0 MaxW Type int Default 0 MaxH Type int Default 0 Minw Type int Default 0 64 MinH Type int Default 0 Range Type int Default 0 Values Auto 0 User 1
439. t is not a true inverse in that when it is applied to a tranformed image it does not exactly reproduce the original 100 image The cHough operator in the reverse mode takes a vector of unsigned 8 or 16 bit images or a single image which is treated as a one element image vector as input and produces a 16 bit output image Each pixel in the input images contributes to a circle in the output image The center of the circle corresponds to the x y position of the input pixel The radius is derived from the element number of the image in the input vector with the first image corresponding to circles with radius startRadius and subsequent images radii increasing in one pixel steps The accumulate parameter determines how the output pixel values are determined from the input image pixel values It may be set to max sum or incr modes These modes are described in detail in the hough operator documentation Parameter Information direction Type int Default 0 Values forward 0 reverse 1 startRadius Type float Default 0 endRadius Type float Default 0 accumulate Type int Default 0 Values max 0 sum 1 incr 2 C Prototype CorOpRtn cor_cHough CorObj In CorObj Out int direction float startRadius float endRadius int accumulate Headers include WITHOME h wFit h Libraries WITHOME lib wObj lib WITHOME lib wFit lib convexHull Find the convex hull of an image or point set ark E
440. t operator the input image must be a binary image Parameter Information width Type int Default 64 height Type int Default 64 sortByLabel Type int Default 1 Values No 0 Yes 1 239 C Prototype CorOpRtn cor_ocrMakeFont CorImage In CorObj ROI CorObj Font CorOcrFont NewFont int width int height int sortByLabel Headers include WITHOME h wOcr h Libraries WITHOME lib wObj lib WITHOME lib wOcr lib ocrMakeSemiFont Create an OcrFont for SEMI OCR A characters Description The ocrMakeSemiFont operator produces a scaled font based on the SEMI standard The operator s output is an OcrFont type object that can be used by a ocrGrayInit or ocrPerimInit operator to create a font model for use with the oerGrayRun or ocrPerimRun operator The operator can produce gray scale font images for use with the ocrGrayInit operator or binary font images for use with the ocrPerimInit operator Parameters style The style parameter is used to specify whether the characters produced are made up of solid lines or a 5x9 dot matrix width and height These parameters specify the size of the characters produced The width and height are measured in pixels from center 240 to center of the strokes or dots that make up the characters clearZone The clearZone parameter is used to specify the size of the empty space between characters in the font strokeWidth The strokeWidth parameter speci
441. tFind2D operator locates the best match or matches to patterns from a 2D array of pattern variations The inputs to the operator are a gray scale target image a rectangle graphic specifying the ROI and a vector of pattern models which may use tiled pattern images and or binary masks A rectangular region of interest ROI of the target image is searched for the pattern matches Matches between the patterns and the image are evaluated by a Statistical technique using pixel grayscale values A two level strategy is used to speed up the search The operator outputs the location match score and pattern index of the match or matches meeting criteria specified by the operator s parameters Inputs Target The target image can be any gray scale WiT image an unsigned or signed eight or sixteen bit integer image or a float image The operator does not accept color or complex images ROI The ROI must be a rectangle type Graphic object The operator searches locations in the target image for which the pattern will be completely contained within the ROI The ROI must be large enough to contain the pattern An ROI specifying the entire 252 Outputs image is produced by the getRect operator PatModel The vector of pattern models is produced by the patInit2D1 operator from from a vector of image vectors The input pattern models in either dimension may represent variations on a pattern or different basic patterns A typical example is
442. ta operator The input vector of Line objects is generally produced by the geom2lines operator Since multiple graphic objects may used to represent each line a single line input results in a vector of graphic objects output The graphic representation can include the lines alone the endpoints of the lines or both as specified by the graphicType parameter The color of the lines and endpoints can be specified separately The color of the output graphics can be controlled using parameters on the Color subpanel The linePen and pointPen parameters are used to turn on and off the pens for lines and points respectively The lineR lineG and lineB and the pointR pointG and pointB control the respective colors Parameter Information graphicType Type int Default 1 Values bitwise OR lines 1 endpoints 2 linePen Type int Default 1 Values off 0 on 1 lineR Type int Default 255 Values 0 255 lineG Type int Default 255 Values 0 255 lineB Type int Default 0 Values 0 255 pointPen Type int Default 1 Values off 0 on 1 pointR Type int Default 255 Values 0 255 pointG Type int Default 0 Values 0 255 pointB Type int Default 0 Values 0 255 Demo iGraph Operators Blobs GetBlobFeatures igr Operators Blobs GetBlobFeaturesMasked igr Operators Blobs SelectFeatures igr C Prototype CorOpRtn cor_plotLines CorObj Lines CorGraphicVector PlottedLines int
443. tangle type Graphic object The operator searches locations in the target image for which the pattern will be completely contained within the ROI The ROI must be large enough to contain the pattern An ROI specifying the entire image is produced by the getRect operator PatModel The pattern model is produced by the patInit1 operator from a pattern image or vector of images The pattern may be broken into equal sized rectangular tiles and or may incorporate an arbitrary binary mask as described in the patInit1 operator documentation Outputs Locations The match locations are produced as a vector of Real points giving the x and y coordinates of the location in the matching region in the target image corresponding to the center pixel in the pattern image For even dimension width or height pattern images the center pixel location is rounded up to the next integer value Only the best match at each location is reported The matches are sorted in order of descending score Scores The match scores corresponding to the locations in the Locations output vector are produced as a vector of float values Indexes The Indexes output is a vector of integers in which each element is the index number of the pattern in the PatModel that produced the match at the corresponding location in the Locations output vector Parameters scale The scale parameter is used to specify resolution reduction factor for the search It can be set to powers of
444. tated image corresponding to the center point of the input image The background area may extend farther than is necessary to contain the original rotated image when the input image is not square In the clip mode the output image is clipped to the original image size and coordinate system The output image in this mode can be considered a fixed window behind which the input image is rotated about a specified point The center of rotation can be specified as the image center or as a user specified point using the rotateAbout parameter Default center for the graphical object is 0 0 The user specified center of rotation is set with the rotatePt parameter In the clip mode the value of this parameter determines what portion of the image remains in the output image window The effect in the expand mode is more subtle The output image is enlarged to contain the entire rotated image as before and the output image center corresponds approximately to the center pixel of the input image but the pixel specified by the rotatePt parameter corresponds exactly to a pixel in the output image When the antiAliasing parameter is set to yes a linear interpolation method is used to determine pixel values in the rotated image When this parameter is set to no input pixel values are assigned directly to a pixel in the rotated image This method is faster than the anti alias method but pixels are assigned with an accuracy of about 1 pixel rather th
445. te the edge of bar Parameter Information noise Type int Default 5 Demo iGraph Operators Matrix Code39 igr C Prototype CorOpRtn cor_getCode39 CorImage Target CorObj ROI CorObj Decoded int noise Headers include WITHOME h wMatrix h Libraries WITHOME lib wObj lib WITHOME lib wMatrix lib getEAN13 Finds and decodes EAN 13 bar code WEAN Description The getEAN13 operator accepts a binary or grayscale image and one or more rectangular regions of interest ROIs then searches the ROIs for EAN13 barcodes If a barcode is found in a region it is decoded to a string Inputs Target The input image must be an unsigned 8 bit grayscale image with the barcode represented as dark bars on a light background Roi A single ROI may be specified by a rectangular Graphic or Geometry object Multiple ROIs can be specified by a vector of Graphic or Geometry objects Outputs Decoded If the ROI is a single object then the output is a String object If a valid barcode is found the output string contains the characters represented by the barcode If no valid barcode is found an empty string is output If the ROI is a vector of rectangles then the operator seeks for a barcode in each ROI The output is a vector of String objects If a valid barcode is found in a ROI the corresponding element of the output vector contains the decoded string If no valid barcode is found in a ROI the corr
446. ted from blobs ers ee by the webGetBlobs operator using minDeltay Min dista ice along y h axis from y centroid images from the input stream or images produced by the webWaterfall operator In these cases the sumRadius Sum of radial distances first row of image does not in general represent the of each perimeter point first row of the image space in which the Feature from centroid field values are determined The offset between the two image spaces is specified by the yOffset sumRadiusSqrd Sum of radial distances parameter squared maxRadiusSqrd Max radial distance Inputs squared In SET The In input is a vector of Features minRadiusSqrd Min radial distance produced by the getFeatures operator squared The Features are usually derived from maxRadDeltaX Max distance along x blobs produced by the webGetBlobs axis from x centroid at operator max radial point BinaryImage The BinaryImage input must be an unsigned eight or sixteen bit binary or labeled image In a binary image all non perimLength maxRadDeltaY Max distance along y axis from y centroid at max radial point 280 minRadDeltaX Min distance along x axis from x centroid at max radial point minRadDeltaY Min distance along y axis from y centroid at max radial point Max length of perimeter Min length of perimeter Max width of perimeter maxLength minLength max Width may minWidth Min
447. ten Vector sinenion 161 getWarpTransform sssssessesseneereerersere 125 DO OEE PAE E E E A hevere 161 insertGraphic F X MICH SIZE naanin en anna anaa MOVE AO EENE EE E eva sapaaetas assess pan polar ae getTiles rotate 129 Bfrcwsccbcecneeeies 165 scroll 131 partition 165 shearX 131 Df vsvvctecscovves 166 shearY a putField 166 WALD veseceeseeee ay renumber 168 zoomScale 132 selectElem 168 ZOOMSIZO onn nman a ann SCLROD aa 169 Measurements esiieisir riisiin 39 shuffle VecEXp iiir iriri ir 169 CAPP rii aitei tE SEE REELE E shuffle Vectoria nonnii a 170 distAllPtToPt or sortObj E CIStMGD isi Eer EPEE E E RRT ii SOTETWOF ICIS ccceeccsceseeseeeeessceseeseceeeeseenseeseees 172 diStPtTOLINe ici ee eee eas ee SWAD BYLES Anin E E EA OE 172 getAngle he tilelmages BOLD ISE es ieee eee eaa indescesedeisbindoulonnses 138 vecRedim Sr DISIMGD ir anand RARE 138 Pixel Access plotAngle see DIOLDISH innnan a a CT ODE AEE EE O S EN E hitOrMiss re Wate soon secs cae nee iinet TAT MARO ENEE AE AEE AAE cht OS LE tN eg median 146 putCol morphClose 147 putPixel morphGradient 148 putRow re morphOpen 149 putVector 180 outline 150 rasterize 180 TANK 151 sampleLine 182 SK CLCLON seasons beeen gute a
448. ten into three separate files corresponding to the red green and blue color channels The three files are suffixed with the extension r g and b Parameter Information filename Type CorFile Default sample wit type Type int Default 7 Values GIF obsolete 0 PCX 1 PBM 2 XBM 3 TIFF 4 BMP 5 JPEG 6 WiT 7 Raw 8 JPEG2000 JP2 9 JPEG2000 CodeStream 10 dataStorage Type int Default 0 Values raw 0 ascii 1 bitmap Type int Default 0 Values no 0 yes 1 quality Type int Default 75 Values 0 100 witFormat Type int Default 0 Values binary 0 text 1 none 2 display 3 colorFormat Type int Default 0 Values packed 0 banded 1 padding Type int Default 2 Values none 0 prepend 1 append 2 numberFiles Type int Default 0 Values 3 0 1 1 order Type int Default 1 Values RGB 0 BGR 1 215 C Prototype CorOpRtn cor_writeImage CorImage In CorFile filename int type int dataStorage int bitmap int quality int witFormat int colorFormat int padding int numberFiles int order Headers include WITHOME h wSystem h Libraries WITHOME lib wObj lib WITHOME lib wSystem lib writeObj Write a WiT object to disk Description The writeObj operator writes any WiT object to a file specified by the filename parameter The current directory is used as the directory to write the file unless an absolute path name is given File
449. ters to the output string until encountered If next character is it is copied Otherwise the following character s up to and including the first conversion character are treated as a printf style format specifier and the string representation of the corresponding input object is appended to the output string Input objects are matched from top Argl to bottom Arg3 with the format specifiers as they are encountered Unneeded extra input objects are just ignored The unused inputs must be connected and must receive an object for the operator to fire An error occurs if the input object type does not match the corresponding format specifier or if more than three arguments are required by the format string The sprint operator produces two outputs Outstr the completed output string and Count the number of chars written to the string The C sprintf conversion characters are d i signed decimal notation o unsigned octal notation x X unsigned hexadecimal notation u unsigned decimal notation c single character s string f fixed point decimal notation e E decimal notation in mantissa exponent form e g G use e or E if exponent is less than 4 otherwise use f p not implemented in WiT n not implemented in WiT Departures from the behaviour of Microsoft s C standard library sprintf e Sprint does not allow the precision to be greater than 510 MSC sprintf may crash Sprint do
450. tes for each pixel For float and complex images the operator reverses the order of the four bytes for each pixel C Prototype CorOpRtn cor_swapBytes CorImage In CorImage Out Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib tilelmages Display a vector of images as one image mi aif Description The tilelmages operator converts a vector of images into one image The resulting image is a grid of images having a number of columns and rows as denoted by the tileColumns and tileRows parameters If either parameter is 0 then the operator determines the width and or height of the grid image The spacing parameter allows the user to place a band of zero valued pixels between each image The number of pixels in that band is given by the value of the parameter Additionally images from the vector can be placed in the top left corner of each grid area or in the center of the area This may be useful when images in the vector are of varying sizes The tiled image size and spacing are calculated using the full size of the input images even if the image ROIs are smaller than the images Only the pixel values from within the input image ROIs are copied to the tiled image This operator is convenient for viewing vectors of related images for example a series of slices through a 3D volume Parameter Information tileColumns Type int Default 0 tileRows Ty
451. tes squared sumysq Sum of all y coordinates squared sumxy Sum of all x coordinates y coordinates perimx X coordinate of a pixel on the blob perimeter lower right corner perimy Y coordinate of a pixel on the blob perimeter lower right corner xc X coordinate of blob centroid sumx n ye Y coordinate of blob centroid sumy n clockwise from x axis major Length of best fit ellipse major axis minor Length of best fit ellipse minor axis xdiff xmax xmintL o lydiff ymax ymins1 o boxarea xdiff ydiff bxaratio npixels boxarea perim Perimeter data for blob optional The perimeter fields are defined as follows perimLength Total length of perimeter of points around object maxDeltaX Max distance along x axis from x centroid minDeltaX Min distance along x axis from x centroid maxDeltaY Max distance along y axis from y centroid minDeltaY Min distance along y axis from y centroid sumRadius Sum of radial distances of each Sawai arco spams point from centroid sumRadiusSqrd_ Sum of radial distances Sum of radial distances squared a Max radial distance a a minRadiusSqrd Min radial distance Min radial distance squared maxRadDeltaX Max distance along x axis from x centroid at max radial point maxRadDeltaY Max distance along y axis from y centroid at max radial point minRadDeltaX Min distance along x axis from x centroid at max radial point minRadDeltaY Min distance alon
452. th and height and the ROI includes the entire image If the parameter is a Geometry or Graphic object or a vector of Geometry Graphic or Blob objects the image and ROI will be the smallest that can contain the object s It the parameter is an image the output image size and ROI will match the size and ROI of the parameter image If the parameter is a vector of images the output image 7 and ROI will be the smallest size large enough to contain all of the images in the input image vector The imageSpec parameter is often used in the input mode with the original image from which the blobs were obtained fed into it to maintain the original image specifications Parameter Information pixelValue Type int Default 0 Values Binary 0 Id 1 Label 2 mode Type int Default 0 Values Fill 0 Outline 1 outType Type int Default 0 Values 8 bit 0 16 bit 1 imageSpec Type CorObj Default OBJ_B T CorPoint 256 256 Demo iGraph Operators Blobs GetBlobFeatures Operators Blobs GetBlobFeaturesMasked C Prototype CorOpRtn cor_labelBlobs CorBlobVector Blobs CorImage Out int pixelValue int mode int outType CorObj imageSpec Headers include WITHOME h wBlob h Libraries WITHOME lib wObj lib WITHOME lib wBlob lib mergeFeatures Merge Blob features into a single set of features Description The mergeFeatures takes a Blob feature vector produced by the getFeatures operator and calculates
453. that a Burt pyramid is a sequence of n images I i the process is to expand I n then add I n 1 The result is expanded and then I n 2 is added etc The filter parameter selects the type of gaussian filter to be applied The effect of each filter is described in the WiT filter library during the discussion of the gauss operator Care must be taken to insure that the filter used is the same one that was used to generated the pyramid or reconstruction will not be exact Parameter Information filter Type int Default 0 Values None 0 2x2 1 3x3 2 5x5 3 7x7 4 final Type int Default 0 start Type int Default 1 C Prototype CorOpRtn cor_burtRecon CorImageVector Pyramid CorImage Out int filter int final int start Headers include WITHOME h wPyramid h Libraries WITHOME lib wObj lib WITHOME lib wPyramid lib 184 burtRecStep One level reconstruction of a Burt pyramid Description The burtRecStep operator performs one iteration in the Burt reconstruction process The Gaussian port input image is the gaussian image for the previous level while the Laplacian port input image is the laplacian difference image for the current level The output is the gaussian image for the current level The filter parameter selects the type of gaussian filter to be applied The effect of each filter is described in the WiT filter library during the discussion of the gauss operator Care must be t
454. the MapValue mode depending on available hardware Parameter Information outType Type int Default 3 Values int8 0 int16 1 int32 2 uint8 3 uintl6 4 uint32 5 float32 6 float64 7 mode Type int Default 0 Values Clip 0 MapRange 1 MapValue 2 MapMSB 3 min Type float Default 0 max Type float Default 1 C Prototype CorOpRtn cor_castImage_1 CorImage In CorImage Out int outType int mode float min float max CorOpRtn cor_castImage_1_consume CorImage In int outType int mode float min float max Headers include WITHOME h wConvert h Libraries WITHOME lib wObj lib WITHOME lib wConvert lib ct Convert basic object to another type atl 42 Description The ct cast operator converts its input object to the output type specified by the type parameter The operator is primarily used to convert between the various integer and float object types including conversions between Point and Real Point objects or to convert numbers to strings or strings to numbers It can be used to cast single objects or vectors of objects Conversions between floating point objects float and double types integer objects byte ubyte short ushort int uint hex8 hex16 hex and char types and between Point and Real Point objects result in the value of the input object being assigned to an output object of the selected type If the value of the input object falls outsi
455. the geom text data field of a graphic object is accessed If the input graphic is not a text graphic then it will generally have an empty geom text field and a text string cannot be inserted directly into the geom text data field The putField operator can also be used to insert values into vectors of complex objects To specify this mode the field parameter is prefixed with a There are three cases to which this may be applied depending on the nature of the input values If the input to be modified is a vector of complex objects and the field value input is a vector of the field type then the elements of the field value vector are inserted into corresponding elements of the complex object vector In this case the two vectors must have the same number of elements If the input to be modified is a vector of complex objects and the field value is a single object of the field type then a copy of that object is inserted into the specified field in each of the complex objects in the vector If the input to be modified is a single complex object and the field value input is a vector of the field type then a vector of copies of the complex object is created and the field values are inserted into the corresponding elements in the vector Parameter Information field Type CorString Default x C Prototype CorOpRtn cor_putField CorObj In CorObj FieldValue CorObj Result CorString field Headers include WITHOME h wObjm
456. the getData image The operator then outputs a vector containing the Graphic objects specified by the user You can also press the Enter key on the keyboard to achieve the same effect Alternately the user can select Cancel in which case execution will terminate with no output produced or Reset which will delete all of the graphical objects and allow the user to start again Operators following getData in the igraph will not be executed until the user defines the data required The type parameter specifies the graphical object type initially desired The choices are points lines rectangles circles polylines polygons or text The object type can be changed interactively by selecting one of the buttons on the lower row of the editor panel Once the object type is selected the object can be drawn on the image using the left mouse button A single click specifies a point or the starting cursor position for text entry Lines rectangles and circles are specified by pressing dragging and releasing the mouse button with the press and release points specifying the two end points of a line the upper left and lower right corners of a rectangle or the center and radius of a circle Polylines and polygons 56 are drawn with a mouse click for each segment end point and a double click to terminate point placement Polyline and polygon points can be deleted in reverse order before the the point placement is terminated by clicking the
457. the length of the arc The number of sample lines is specified by the numLines parameter The length of the radial sample lines is specified by the lineLength parameter The radiial sample lines are centered on the nominal circle if its radius is greater than half the sample line length otherwise the lines extend from the nominal circle center out to the specified length Image value sampling along the sample line is controlled by the line Width and smoothing parameters as described for the getXEdgesOnLine operator Edge selection is controlled by the edgeType slopeThresh levelThresh and subpixel parameters These parameters are also explained in the getXEdgesOnLine operator help If the edgeType parameter is set to unknown the operator will chose either all rising or all falling edges based on which type has the greater average level change over all the sample lines 73 Parameter Information slopeThresh Type float levelThresh Type float edgeType Type int Default 0 Values rising 0 falling 1 unknown 2 subpixel Type int Default 0 Values No 0 Yes 1 lineLength Type float Default 20 numLines Type int Default 12 lineWidth Type int Default 1 smoothing Type int Default 1 Demo iGraph Operators Edges FindCircle C Prototype CorOpRtn cor_getCircularXEdges CorImage In CorObj NomCircle CorEdgeVector PtEdges CorEdge CircleEdge int NFound float slopeThresh float levelThresh
458. the minor axis of the ellipse typically called b If the y coordinate is set to zero the underlying ellipse is a circle with the radius specified by the x coordinate of this point The x coordinate of the fourth point is the angle of the major axis of the ellipse clockwise from the x axis in degrees The x coordinate of the third point represents the angle of the starting point of the arc on the ellipse with respect to the major axis The ycoordinate represents the angle subtended by the arc These angles values represent the angles after normalizing the ellipse to a circle An If the number of rows in the points array is less than the number of points required by the graphic object the missing points will be set to 0 0 If the number of rows is greater than the number of points needed the extra points will be ignored The color of the output graphics can be controlled using parameters on the Color subpanel The drawing method can be specified using the outline and fill parameters The outline and fill colors can be chosen with the penR penG and penB parameters and fillR fillG and fillB parameters If fill alone is chosen the filled object includes the outline If both outline and fill are chosen the outline color is written over the fill color on the object outline When Text is selected a text string can be entered using the text parameter of the Text subpanel The font characteristics for the text can be set using the fon
459. the output image is set to the parameter value up to a maximum of the corresponding dimension in the input image When a parameter is set to zero the corresponding output dimension is the same size as the dimension in the input image A reduced size image is centered on the original image The width and height parameters may be used to reduce the size of the rotated output images so that the corners of the image contain valid image data When regions from outside the input image are rotated into the field of view of an output image the pixels are set to the average of the four corner pixels of the input image Parameter Information minAngle Type float Default 0 maxAngle Type float Default 0 angleIncr Type float Default 1 Type int Default 0 Type int Default 0 CorImageVector Rotated CorFloatVector Params float minAngle float maxAngle float anglelIncr int width int height Headers include WITHOME h wSearch h Libraries WITHOME lib wObj lib WITHOME lib wSearch lib patScale Rescale a pattern image through a range of sizes Mg Description The patScale operator creates a vector of scaled versions of the input image The input image is scaled in increments specified by the operator s parameters to produce the images in the output image vector This operator can be used to produce a vector of images for input into the patInit operator The resulting pattern model v
460. then refined at full and possibly subpixel resolution The parameters on the Candidates subpanel are used to control the low resolution search and candidate selection scale The scale parameter is used to specify resolution reduction factor for the initial search for candidates It can be set to powers of two from two to 32 max Candidates The maxCandidates parameter specifies the maximum total number of candidates that will be considered in the first phase search This maximum limit is applied before the candsInVicinity and candRelThresh parameters which can further reduce the number of candidates passed to the second search phase candsInVicinity The candsInVicinity parameter specifies the maximum number of candidates that are passed to the second phase from within any subregion of the image of a size specified by the xVicinity and yVicinity parameters candThresh The candThresh parameter is used to set the rejection threshold for candidate matches candRelThresh The candRelThresh parameter is used to set a relative rejection threshold for candidate matches The relative rejection threshold is determined by multiplying the parameter value by the match score of the best candidate match found The relative threshold is used to restrict the number of candidates passed on to the second to those near the best in each region of the image output The Candidates setting of of the output parameter allows the user to output t
461. tination position for the goto operator The instance name of label is used to match the destination specified in the label parameter of goto Demo iGraph Operators clearObjs memConst Store constant objects for later recall Description Store constant objects for later recall Parameter Information name Type CorString Default mem value Type CorObj 119 persist Type int Default 0 Values no 0 yes 1 Demo iGraph Operators clearObjs memFree Free previously stored objects Ca J Description Free memory previously allocated by a memStore operator Memory stored using memStore is usually freed automatically when a new igraph is loaded Calling memFree explicitly is useful when an object is no longer needed and it takes up too much memory Objects stored in memory are referenced by the name parameter which refers to a global name See chapter Operators Special in User s Manual for details See also memStore memRecall Parameter Information name Type CorString Default mem Demo iGraph Operators clearObjs memRecall Recall previously stored objects 120 Description Recall memory previously allocated by a memStore operator Objects stored in memory are referenced by the name parameter which refers to a global name See chapter Operators Special in User s Manual for details See also memStore memFree Parameter Information name Type CorStrin
462. ting pattern model vector can then be used to search for rotated patterns using the patFind operator Inputs In The input image can be a gray scale or color image complex images are not currently supported by this operator Outputs Rotated The output is a vector of rotated versions of the input image The number of rotated images the rotation angles and the image sizes are controlled by the parameters as described below The images in the output vector are of the same type as the input image Note that vectors color images produced by this operator cannot be used as input to the patInit operator Params The Params output is a four element float vector that contains the transformation parameters applied to the images in output vector The first element is the minimum scale factor the second is the scale increment the third is the minimum 265 rotation angle and the fourth is the angle increment For the patRotate operator the first value is always 1 0 and the second is always 0 Parameters startAngle endAngle and angleIncr The input image is rotated by angles from startAngle up to the largest angle less than or equal to endAngle in increments of angleIncr The value of startAngle must be less than or equal to the value of endAngle and angleIncr must zero width and height The width and height parameters control the size of the output image When a parameter is set to a non zero value the corresponding dimension of
463. tion The acqFrameTrigger operator is used to control the trigger state of the currently selected frame grabber When the source is set to Internal the acquire operator will grab an image from the connected camera based on triggering from an internal generator on the frame grabber board If the source is External the acquire operator will block until the frame grabber board receives an external trigger input to start a grab If the source is None the board will send no frame trigger signals to the camera Parameter Information source Type int Default 1 Values None 0 Internal 1 External 2 C Prototype CorOpRtn acqFrameTrigger int source Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqHue Set video hue and saturation on frame grabber 293 Description The acqHue operator is available for frame grabbers that support color acquisition This operator is used to set the hue and saturation used by the acquire operator The range of these parameters is expressed as a percentage between 0 and 100 Parameter Information hue Type int Default 50 Values 0 100 saturation Type int Default 50 Values 0 100 C Prototype CorOpRtn acqHue int hue int saturation Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqloRead Read from external I O on frame grabber Descrip
464. tion The acqloRead operator is used to read a digital T O lines from a selected port which may be available on the frame grabber card 294 The port parameter is used to select which port to read from The operator s output is an integer which can be interpreted as a bit field where a bit value of 1 indicates that the corresponding line is on or high and a value of 0 indicates that the line is off or low See the description of the acqIloSetup operator for more information on the use of hardware I O Parameter Information port Type uint Default 1 C Prototype CorOpRtn acqIoRead uint Bits uint port Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqloRelease Release external I O resources on frame grabber a at Description The acqloRelease operator is used to release the hardware resources used to read or write I O lines on a frame grabber See the help for acqIoSetup for more information on the use of frame grabber I O lines Parameter Information port Type uint Default 0 C Prototype CorOpRtn acqIoRelease uint port Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib acqloSetup Configure external I O on frame grabber Description The acqloSetup operator is used to configure the hardware resources necessary to read and or write to I O lines available on a frame
465. tion parameter determines whether a minimum and or maximum width height and number of pixels will be specified for blobs extracted from the image ROIs When Min size is selected the minimum values are set using the minWidth minHeight and minPixels parameters When Max size is selected the maximum values are set using the max Width maxHeight and maxPixels parameters A blob must meet all the criteria to be selected Blobs outside the ranges are always ignored Blob selection is not affected by the ignoreNoMatch parameter setting Blobs that do not meet the selection criteria are not compared to the model font characters so selection is usually a more efficient way of eliminating extraneous small or large blobs than using a limited scale range minWidth and minHeight These parameters specify the minimum blob width and height in pixels when the Min size choice of the selection parameter is activated max Width and maxHeight These parameters specify the maximum blob width and height in pixels when the Max size choice of the selection parameter is activated Parameter Information rotation Type int Default 0 Values Invariant 0 Non Invariant 1 Range 2 sensitivity Type float Default 0 5 minAngle Type float Default 5 maxAngle Type float Default 5 featuresUsed Type float Default 0 5 scaleInvariant Type int Default 1 Values No 0 Yes 1 minScale Type float Default 0 5 maxScale Type float Def
466. tor searches for edges along a number of radial sample lines through the image that are defined by the NomCircle and the numLines parameter The operator takes the best edge found on each radial sample line that meets the criteria specified in the operator s parameters and fits a circle to them using a least squares fit The operator s outputs are PtEdges a vector of point type edges containing the edge point found on each radial sample line CircleEdge an Edge object containing the best fit circle or arc and NFound the number of edge points found The CircleEdge output will be a circle if the NomCircle input is a circle and an arc if NomCircle is an arc The slope and level field values of the CircleKdge are the average of the corresponding edge point values Positive slope values represent rising edges and negative slope values represent falling edges For circles the direction field value is always zero and for arcs the direction field value is the angle of the radial line to the center of the arc It should be noted that the operator will output a circle even when fewer than three edge points are found If no edge points are found the operator outputs the nominal circle If one or two edge points are found the operator outputs the nominal circle translated to include the edge point or points These conditions can be detected by monitoring the NFound output value The radial sample lines are spaced evenly around the circle or over
467. totype CorOpRtn cor_getPOSTNET Headers CorImage Target CorObj ROI CorObj Decoded int maxSize int timeout include WITHOME h wMatrix h Libraries WITHOME lib wObj lib WITHOME lib wMatrix lib getPharmacode Finds and decodes Code 39 bar code Description The getPharmacode operator accepts a binary or grayscale image and one or more rectangular regions of interest ROIs then searches the ROIs for Pharmacode barcodes If a barcode is found in a region it is decoded to a string Inputs Outputs Target The input image must be an unsigned 8 bit grayscale image with the barcode represented as dark bars on a light background Roi A single ROI may be specified by a rectangular Graphic or Geometry object Multiple ROIs can be specified by a vector of Graphic or Geometry objects ROI must include silence zones on both sides of the barcode Decoded If the ROI is a single object then the output is a String object If a valid barcode is found the output string contains the characters represented by the barcode including the check character If no valid barcode is found an empty string is output If the ROI is a vector of rectangles then the operator seeks for a barcode in each ROI The output is a vector of String objects If a valid barcode is found in a ROI the corresponding element of the output vector contains the decoded string If no valid barcode is found in a ROI the corresp
468. tructuring element or neighbourhood specifying kernel used The structuring element may be specified as a rectangular region with the size of the region is given by the width and height parameters or it may be specified as an arbitrary 147 region in which case the neighbourhood is specified using the kernel parameter as described in the erode or dilate documentation Closing by a rectangular structuring element is usually faster than closing by an arbitrary element particularly for larger neighbourhoods One common exception is closing by 3x3 structuring elements where the arbitrary kernel usually is faster Parameter Information type Type int Default 0 Values Binary 0 Grayscale 1 kernelType Type int Default 0 Values Rectangular 0 Arbitrary 1 width Type int Default 7 height Type int Default 7 kernel Type Corlmage Default 1 1 1 1 1 1 1 1 1 optimization Type int Default 1 Values preciseness 0 speed 1 Demo iGraph Operators Morphology hitormiss igr C Prototype CorOpRtn cor_morphClose CorImage In CorImage Out int type int kernelType int width int height CorIntImage kernel int optimization CorOpRtn cor_morphClose_consume CorImage In int type int kernelType int width int height CorIntImage kernel int optimization 148 Headers include WITHOME h wMorpho h Libraries WITHOME lib wObj lib WITHOME lib wMorpho lib morphGradient
469. ts the first edge of the input polygon If one edge point is found the operator outputs a line parallel to the first edge of the polygon that goes through the single edge point These conditions can be detected by monitoring the NFound output value The input polygon should be a quadrilateral or triangle The starting end points of the sample lines are equally spaced along the first edge of the polygon If the polygon is a triangle the third vertex point is the second end point for all of the sample lines If the the polygon has four or more sides the sample lines second endpoints are equally spaced along the third edge of the polygon The number of sample lines is specified by the numLines parameter Image value sampling along the sample line is controlled by the lineWidth and smoothing parameters as described for the getXEdgesOnLine operator Edge selection is controlled by the edgeType slopeThresh levelThresh and subpixel parameters These parameters are also explained in the getXEdgesOnLine operator help If the edgeType parameter is set to unknown the operator will chose either all rising or all falling edges based on which type has the greater average level change over all the sample lines The LineEdge slope value is positive if the edge is a rising edge and negative if the edge is a falling edge Parameter Information slopeThresh Type float levelThresh Type float edgeType Type int Default 0 Values rising 0 f
470. ttern will be completely contained within the ROI The ROI must be large enough to contain the pattern An ROI specifying the entire image is produced by the getRect operator PatModel The pattern model is produced by the patInit1 operator from a pattern image or vector of images The pattern may be broken into equal sized rectangular tiles and or may incorporate an arbitrary binary mask as described in the patInit1 operator documentation Outputs Locations The match locations are produced as a vector of Real points giving the x and y coordinates of the location in the matching region in the target image corresponding to the center pixel in the pattern image For even dimension width or height pattern images the center pixel location is rounded up to the next integer value Only the best match at each location is reported The matches are sorted in order of descending score Scores The match scores corresponding to the locations in the Locations output vector are produced as a vector of float values Indexes The Indexes output is a vector of integers in which each element is the index number of the pattern in the PatModel that produced the match at the corresponding location in the Locations output vector Parameters rejectThresh The rejectThresh parameter is used to set a rejection threshold for match scores below which no matches will be considered The parameter must be set to a value from zero to one where zero
471. tures operator to an image formed by concatenating the images in the stream into a single image The difference is that this operator provides parameters to reset the operator s state and to limit the maximum height of the blobs produced and perimeters cannot be extracted by this operator Inputs BinaryImage The BinaryImage input must be an 8 or 16 bit unsigned image The images may be interpreted as either binary images such as those produced by the thresh operator or labeled images such as those produced by the kMeans operator All images received by an instance of the operator must be the same size and type Outputs BlobFeatures The BlobFeatures output is a Blob feature vector Each element of the vector contains information related to one contiguous region in the input image This information includes such things as the number of pixels in the region and its centroid The Blob feature structure comprises the following fields Field Description Name id Unique blob identifier npixels Number of pixels in the blob xmin Minimum x coordinate of blob left edge xmax Maximum x coordinate of blob right edge ymin Minimum y coordinate of blob top edge ymax Maximum y coordinate of blob bottom edge sumx Sum of all x coordinates sumy Sum of all y coordinates sumxsq_ Sum of all x coordinates squared sumysq Sum of all y coordinates squared sumxy Sum of all x coordinates y coordinates perimx X c
472. turns the maximum and minimum values of the pixels or vector elements Bounding values can be found for integer float and complex images but not for color RGB images For a complex image the maximum and minimum of the base 10 logarithm of the magnitude of the complex pixel values are produced Vector bounding can only be performed on vectors of simple objects Bounding values for vectors of vectors or vectors of images cannot be found using this operator Demo iGraph Operators Statistics BoundThresh C Prototype CorOpRtn cor_bound CorObj In double Min double Max Headers include WITHOME h wStats h Libraries WITHOME lib wObj lib WITHOME lib wStats lib countPix Count the non zero pixels in an image Description The countPix operator counts the non zero pixels in the input image Demo iGraph Operators Statistics CountPix C Prototype CorOpRtn cor_countPix CorImage In int Count Headers include WITHOME h wStats h Libraries WITHOME lib wObj lib WITHOME lib wStats lib difference Measure the difference between two vectors or images a a a ABIE pepeg Description The difference operator computes the difference between two images or vectors of numbers in one of several of ways The various measures are based on pixel by pixel or element by element differences in the image or vector The measure parameter specifies how the pixel or element value dif
473. tween banded and packed color formats 35 Description The convertColorFormat operator converts between the obsolete WIT_RGB banded color image format and WIT_COLOR packed color image format This operator is intended primarily to maintain compatability for user written operators that use the WIT_RGB format Existing WiT operators that accept color images as input will process WIT_RGB or WIT_COLOR format images Operators that produce color images as output will produce WIT_COLOR format images In the future new operators or modified operators that can accept color images may not accept the old format The format change will not affect the operation of graphs that use only existing WiT operators however it may be necessary to add a convertColorFormat operator to existing graphs convert color images to the banded format if they are input to user written operators In addition it may be necessary to convert the output from user written operators or older stored images to the packed format if they are input to new WiT operators or operators that previously did not handle color input This operator is intended to facilitate transition to the new WIT_COLOR format and we recommend that operators be converted to use the new format Parameter Information direction Type int Default 0 Values gt packed 0 gt banded 1 Demo iGraph Operators Color HS V Operators Color Brighten C Prototype CorOpRtn cor_convertColorFormat
474. ty output vectors are produced acceptThresh The acceptThresh parameter specifies an acceptance threshold above which a match score value is always considered a valid match If the maximum number of matches specified by the maxMatches parameter are found with match score above the acceptThresh value the search stops immediately This parameter is used to control early termination of the search for improved speed minContrast The minContrast parameter is used to specify the minimum contrast required in the target image to produce a match The parameter specifies the minimum standard deviation of the pixel values in a possible match region relative to the standard deviation of pixel values in the pattern model image This parameter is usually set to a value between zero and one although values greater than one can be used if the target image has higher contrast than the model images max TilesOmitted If the pattern model breaks the pattern images into multiple tiles the maxTilesOmitted parameter allows the operator to omit some of the tiles while calculating the best matching score The operator will use the best score for a location that omits at most the number of tiles specified by this parameter The operator will always use at least one tile to calculate the best score Setting this parameter to a value greater than zero provides some tolerance to partially occluded or damaged target regions localSearch The localSearch par
475. type Type int Default 1 Values Binary 0 Grayscale 1 kernelType Type int Default 1 Values Rectangular 0 Arbitrary 1 width Type int Default 7 height Type int Default 7 kernel Type Corlmage Default 1 1 1 1 1 1 1 1 1 optimization Type int Default 1 Values preciseness 0 speed 1 Demo iGraph Operators Morphology hitormiss igr C Prototype CorOpRtn cor_morphGradient CorImage In CorImage Out int type int kernelType int width int height CoriIntImage kernel int optimization CorOpRtn cor_morphGradient_consume CorImage In int type int kernelType int width int height CorIntImage kernel int optimization Headers include WITHOME h wMorpho h Libraries WITHOME lib wObj lib WITHOME lib wMorpho lib morphOpen Morphological opening Description The morphOpen operator performs binary or gray scale morphological opening on images Binary opening tends to remove small isolated pixels or small regions extending from larger objects without making the larger objects smaller Binary opening is implemented by applying binary erosion followed by binary dilation to the image Gray scale opening tends to reduce pixel values in small high valued regions without reducing the value at the edges of large high valued regions Gray scale opening is implemented by applying gray scale erosion followed by gray scale dilation to the image This operator could be
476. uPs 0 3h 74 Cont tosis eine weet ae I 112 getXEdges 74 n EAE cia a sheet eaten ak cS aot 112 getXEdgesInPoly 76 COO eth h hic a ee thi eae 112 getXEdgesOnLine se Pl CLAP OD Snan ean aa eta 112 trace AGES iis cveveesneeh cchsceng hte ai 78 COMCCIOP denii A he 112 Bilt E EEEE compass CONV Gass sasesrsvnscuseaiessosdeags cases A AE ERRA Ei CONVZE ede aa R A EE RS edgeDetect a CAGCDILECHION cesceeecsecsseseeseeseeseesenseteetereeeneesees 82 1A AA AEE ANA E tek EEA EEAS 115 CNL OPY ERER AO EAEE Ea R GAUSS nanana lopassld eS LOPASS2 d akena ete doce ear tte ee eed PEW Oh coli tei ockeh aaae e a ae eeaeee refineEdges 90 ifCond3 00 ee SOpel asisas lt 9 ifConditional thinEdges 91 LN EEEE AS variance 92 label Find 92 memcConst getMatch ie 92 memFre getPeaks 94 memRecall NXCOTT 95 memStore SOli 95 MUXS eeeeeeeees miath eneee a e E E eee es 96 MUN renane E EE E EAE N tmatchSparsen esn e e QS oneshot srn aaor rE EEE EERE xCorr Ut su iat Sas teaches dai cabsdesnand le deishiesecadeddhcel AE a AE E ies EEE E convexHull PUC TeleTOP tS i neen shee eaten 101 SEAT ES AAIE NAE T et PULINCT OPIS dennes ienasi inea a ke 102 stopDisable asoan iinei iiaii fitTwoLinesToPts stopEnable Centroid i oi 5 sae eee EEE aas VOLE EE EE gFitContoursRoi osese GOME Y a e ear E E E RE getRigidBodyTransform eesse 124 flat
477. ues round 0 floor 1 ceiling 2 C Prototype CorOpRtn cor_cast CorObj In CorObj Out int type int conversion Headers include WITHOME h wConvert h Libraries WITHOME lib wObj lib WITHOME lib wConvert lib castimage Convert image from one type to another a4 i Description The castImage operator converts an integer or floating point input image to a specified integer or floating point type The input image can be any scalar image type int8 int16 int32 uint8 uintl6 uint32 float32 or float64 The output image may be any scalar image type as specified by the outType parameter WiT processing and display operators do not normally handle int32 uint32 or float64 format images Images in these formats can be created using for example the crObject operator or vectorsToImage operator Images in the int32 format are used for the kernel parameter in many of the operators in the Filter and Morphology libraries Images in these formats can only be cast in the Clip conversion mode The mode parameter allows the selection of Clip which preserves image pixel values where possible or one of three mapping modes which scale pixel values to fit the output image range In the Clip mode original image pixel values are retained where they fall within the range of the output image type Input image pixel values that fall outside the output image range are set to the range minimum or maximum depending on
478. uplicates When it is set to LocAndPat matches that are within the vicinity and correspond to the same element in the pattern vector are considered duplicates Only the best match of a set of duplicate matches is kept xVicinity and yVicinity These parameters specify a minimum region around a match location in which matches with lower match scores will be suppressed when the location is being used to select matches Two matches will never be produced that are less than both xVicinity pixels apart in x and yVicinity pixels apart in y when the selectMode parameter is set to Location or Both The MultiplePatterns Sub Panel The MultiplePatterns sub panel includes parameters that control the output when multiple pattern models are used applyPatterns The applyPatterns parameter is only meaningful when more than one pattern model is applied When the parameter is set to All all patterns in the input pattern vector are compared to the target image in both phases of the search When the parameter is set to Sample the pattern vector is sampled as specified by the patSampling parameter for the candidate search A range of adjacent pattern models can be selected by setting the parameter to Range in which case the elements used are specified by the patCenter and patRange parameters patSampling The patSampling parameter specifies the sample spacing in the pattern vector for 261 the search when applyPatterns is set to Sample patCenter
479. ure id number will be saved by the operator so that the id of the first Feature produced the next time the operator fires will follow the last id among the current output Features When it is set to Yes the operator will reset the Feature id to zero after the current Features are produced mode The mode parameter setting specifies whether a circle or ellipse will be used to determine if two Features are near each other aScale bScale aOffset bOffset These parameters are used with the best fit ellipse axes to determine the size of the circle or ellipse used In the circle mode the circle s diameter is given by diameter major aScale aOffset In the ellipse mode the ellipse s extended major and minor axes are given by extended major major aScale aOffset extended minor minor bScale bOffset where major is the length of the major axis of the best fit ellipse and minor is the length of its minor axis The circle or ellipse is centered at at the Feature centroid Parameter Information yOffset Type int Default 0 blockHeight Type int Default 32 clear Type int Default 0 resetld Type int Default 0 mode Type int Default 1 Values Circle 0 Ellipse 1 aScale Type float Default 1 bScale Type float Default 1 aOffset Type float Default 0 bOffset Type float Default 0 C Prototype CorOpRtn cor_webClusterFeatures CorFeatureVector In CorFeatureVector Out i
480. ures CorImage SourceImage CorFloatVector Transform CorFeatureVector Warped int connected int type Headers include WITHOME h wBlob h Libraries WITHOME lib wObj lib WITHOME lib wBlob lib labelBlobs Convert blobs to an image E Description The labelBlobs operator converts the input blob vector into an image Blob vectors are produced by the getBlobFeatures operator The mode parameter is used to specify whether the output image contains the filled blob or just the blob outline The blob outline produced by this operator only includes the end points of the blob run lengths rather than a complete outline of the blob The pixel values used for the blobs can be controlled by the pixelValue parameter When this parameter is set to Binary all pixel values in the blob or outline are set to 255 when set to Id the blob id value is used and when set to Label the blob label value is used For Binary values the output image is always an 8 bit unsigned image For Id or Label the outType parameter can be used to specify a 16 bit unsigned output image if the id or label values are expected to exceed 255 The imageSpec parameter is used to specify the size and ROI of the output image This parameter may be a Point Real Point Geometry Graphic or Image object or a vector of Geometry Graphic Image or Blob objects If the parameter is a Point or Real Point type value the x and y values specify the image wid
481. ut is a vector of floats that correspond to the extracted value at each sample location on the line The PO output is a real point that gives the location of the first sample point on the line The Incr output is a real point with the x increment and y increment between sample points in its x and y fields respectively Parameter Information lineWidth Type int Default 1 C Prototype CorOpRtn cor_sampleLine CorImage In CorGraphic Line CorFpointVector Points CorFloatVector Values CorFpoint P0O CorFpoint Incr int lineWidth Headers include WITHOME h wPixel h Libraries WITHOME lib wObj lib WITHOME lib wPixel lib zerolmage Create a blank image Description The zeroImage operator produces an image of the same type and size as the input image with all pixel values set to zero C Prototype CorOpRtn cor_zerolImage CorImage In CorImage Out CorOpRtn cor_zeroImage_consume CorImage In Headers include WITHOME h wPixel h Libraries WITHOME lib wObj lib WITHOME lib wPixel lib Pyramids The Pyramid library provides support for multi resolution image processing The library includes operators that reduce or expand image size operators that produce multi scale pyramids from images and operators that reconstruct images from pyramids burt Generate a pyramid using Burt method ah Ae Description The burt operator generates a vector of n images I
482. ut of the for operator When the currentCount value is equal to count the the input token is sent to the lower branch labelled Break As currentCount is incremented its current value is interactively displayed on the igraph When you run the igraph again the current value of currentCount will revert to zero See also for2 for3 Parameter Information currentCount Type int Default 0 count Type int Default 1 115 Demo iGraph Operators clearObjs for2 Repeat for 2 inputs a set of operations a specified number of times a wane Description Similar to the for operator but controlling 2 objects with the same loop count See also for for3 Parameter Information currentCount Type int Default 0 count Type int Default 1 Demo iGraph Operators clearObjs for3 Repeat for 3 inputs a set of operations a specified number of times Prt eae Description Similar to the for operator but controlling 3 objects with the same loop count 116 See also for for2 Parameter Information currentCount Type int Default 0 count Type int Default 1 Demo iGraph Operators clearObjs gate Gate object data Description The gate operator allows objects to travel from input to output if the Gate input is present otherwise the input object is blocked Both the Gate input and input object must be present before the gate operator executes Demo iGraph Operators clearObj
483. value A gt threshold k 0 thresholding A k if A gt threshold otherwise 0 slicing with background A kifrl lt A lt r2 otherwise A A lt r2 A gt r1 k A A slicing without A lt r2 A gt r1 background k 0 0 A kifrl lt A lt r2 otherwise 0 range compression k log10 1 k log10 1 A c log10 1 abs A abs A where c kK log10 1 k bit removal MSB 2 A k 1 A 2 A k 1 bit removal LSB 2 A 2 A 2 A 2 bit extraction k A 2 8 n A k 2 9 2 A 2 8 n 1 where KSAP j A 2 n 1 B number of bits per image n n most significant bit to extract Parameter Information expression Type CorString 24 Default A userConstant Type int Default 0 inputType Type int Default 0 Values ubyte 0 byte 1 ushort 2 short 3 outputType Type int Default 0 Values ubyte 0 byte 1 ushort 2 short 3 float 4 conversion Type int Default 0 Values round 0 floor 1 ceiling 2 Demo iGraph Operators Calculate Lut C Prototype CorOpRtn cor_makeLut CorVector LUT CorString expression int userConstant int inputType int outputType int conversion Headers include WITHOME h wCalcul h Libraries WITHOME lib wObj lib WITHOMB lib wCalcul lib makeSeq Generate a vector containing a number sequence rca ELETE
484. value a T Description The acquireStamped operator is identical to the acquire operator in all respects except that it also returns the value of the hardware event counter at the moment the frame was acquired The acqCounterReset operator must be executed at least once before acquireStamped executes for the first time in order to initialize the counter 304 Parameter Information frames Type int Default 1 delay Type int Default 0 C Prototype CorOpRtn acquireStamped CorObj Count CorObj Out int Dropped int frames int delay Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib brightness Set video offset and gain on frame grabber Description The brightness operator sets the contrast and brightness of an image acquired with the acquire operator The parameters are given as a percentage between 0 and 100 where 0 is the darkest and 100 is the brightest or greatest amount of contrast Parameter Information brightness Type int Default 50 Values 0 100 gain Type int Default 50 Values 0 100 C Prototype CorOpRtn brightness int brightness int gain Headers include WITHOME h wVideo h Libraries WITHOME lib wObj lib WITHOME lib wVideo lib 305 306 Coreco Imaging Contact Information Sales Information Web site www imaging com Email info corecoimaging com Technical Support Voice
485. value can be used to specify a constant length output vector The following examples give the expression and lengthExpr strings required for some common mappings First K elements if K gt input vector width the last element will be repeated to extend the output vector to K elements expression A lengthExpr K Last K elements expression W K A lengthExpr K Even elements expression A 2 lengthExpr ceil W 2 Odd elements expression A 2 1 lengthExpr floor W 2 All even elements followed by all odd elements 170 expression A lt W 2 A 2 2 A ceil W 2 1 lengthExpr W Parameter Information expression Type CorString Default A lengthExpr Type CorString Default W userConstant Type double C Prototype CorOpRtn cor_shuffleVecExpr CorVector In CorVector Out CorVector Indexes CorString expression CorString lengthExpr double userConstant Headers include WITHOME h wObjman h Libraries WITHOME lib wObj lib WITHOME lib wObjman lib shuffleVector Reorder a vector Description The shuffleVector operator selects and rearranges the elements of an input object vector based on the values of a input index vector of integers The output is a vector made up of the elements from the input object vector at the positions specified in the input index vector The output vector is same type as input object vector and the same size as t
486. various parameters that control candidate selection The MultipleMatches Sub Panel The MultipleMatches sub panel includes parameters that control the output when multiple matches are required max Matches The maxMatches parameter sets the maximum number of matches that will be produced by the operator The operator may produce fewer matches relThresh The relThresh parameter is used to set a relative rejection threshold for the matches The relative rejection threshold is determined by multiplying the parameter value by the match score of the best match found This value is then applied like the normal rejection threshold x Vicinity and yVicinity These parameters specify a minimum region around a match location in which matches with lower match scores will be suppressed Two matches will never be produced that are less than both xVicinity pixels apart in x and yVicinity pixels apart in y Parameter Information rejectThresh Type float Default 0 acceptThresh Type float Default 1 minContrast Type float Default 0 maxTilesOmitted Type int Default 0 localSearch Type int Default 0 Values Fast 0 All 1 subPixel Type int Default 0 Values No 0 Yes 1 resolution Type int Default 3 Values 2 0 4 1 8 2 16 3 32 4 64 5 scale Type int Default 2 Values 2 0 4 1 8 2 16 3 32 4 maxCandidates Type int Default 4 candidateThreshold Type float Default 0 5 candidateRelThr
487. ve integer match non zero pixels in the input image and kernel elements set to 1 or any negative integer match all pixel values in the input image The input and output of the hitOrMiss operator are both 8 bit unsigned images Parameter Information structElem Type Corlmage Default 1 1 1 1 1 1 1 1 1 Demo iGraph Operators Morphology hitormiss igr C Prototype CorOpRtn cor_hitOrMiss CorImage In CorImage Out CorIntImage structElem CorOpRtn cor_hitOrMiss_consume CorImage In CorIntImage structElem Headers include WITHOME h wMorpho h Libraries WITHOME lib wObj lib WITHOME lib wMorpho lib Idilate Labeled image dilation Description The Idilate operator dilates labeled images maintaining image pixel values in the dilated images The effect of this operator is the equivalent to applying the bdilate binary dilation operator to the pixels at each distinct non zero pixel value in the image and assigning the pixel value to the active pixels in the output image except where a pixel would be assigned more than one distinct value in which case it is set to zero The result is that the regions in the image are dilated with equal valued regions growing together and different valued regions remaining distinct Zero valued pixels may or may not be maintained between adjacent regions of different values depending on the size of the kernel structuring element and the distance betwee
488. vector into the input image pixels falling on an arbitrary line The line in the image is specified by the start and end parameters The pixels are selected by rasterizing the line connecting the two end points and consequently do not represent evenly spaced samples Parameter Information start Type CorPoint 180 Default 0 0 end Type CorPoint Default 0 0 C Prototype CorOpRtn cor_putVector CorImage In CorVector Profile CorImage Out CorPoint start CorPoint end Headers include WITHOME h wPixel h Libraries WITHOME lib wObj lib WITHOME lib wPixel lib rasterize Generate an image from graphical objects F ey Description The rasterize operator rasterizes a vector of Graphic objects which describe a set of ROIs in an integer or float image The operator will rasterize geometric objects points lines rectangles polylines polygons and circles but ignores text graphical objects The graphical objects can be drawn in outline or filled or both using different colors in color or grayscale either on a grayscale background image or a black background The overlay parameter determines if the output image is to be zeroed or set to the input image before rasterizing When the parameter is set to off the graphic objects are drawn on a black zero valued background When set to on the image is scaled to the 8 bit integer range 0 255 as specified by the rangeType paramete
489. verse Reverse the threshold range E g if the image is 8 bit unsigned and level is single at 128 then normally pixel values above 128 will be set to 255 pixels below 128 will be set to 0 But if Inverse is chosen then pixels above 128 will be black and pixels below 128 will be white instead Show Enable disable histogram display on histogram ramp Hist max Control scale of histogram displayed on ramp The larger the scale the taller the histogram will appear Default value is maximum value of histogram Ramp Similar to the ramp on the image property panel If Levels is set to Single there is only one pointer on the ramp which indicates a singe threshold level If Levels is set to double two pointers appear on the ramp f Parameter Information name Type CorString Default iThresh invert Type int Default 0 Values No 0 Yes 1 levels Type int Default 0 Values Single 0 Double 1 result Type int Default 0 Values Binary 0 Semi Threshold 1 Placement Type int Default 0 Values bitwise OR Fix position 1 X Type int Default 0 Y Type int Default 0 overlayData Display graphical data overlaid on image Description The overlayData operator provides a means of viewing data generated by getData or other operators with outputs in Graphic format such as plotFeatures The operator accepts an image of any type and a vector of graphic objects as input The inpu
490. ween each component image within the composite image Parameter Information orientation Type int Default 0 Values Horizontal 0 Vertical 1 Spiral 2 spacing Type int Default 0 C Prototype CorOpRtn cor_showPyramid CorImageVector Pyramid CorImage Out int orientation int spacing Headers include WITHOME h wPyramid h Libraries WITHOME lib wObj lib WITHOME lib wPyramid lib unDecimate Expand image without filtering Thais a Miji g r Description The unDecimate operator takes an input image and doubles its size The data pixels from the input image are distributed throughout the output image in one of four possible decimation lattices The lattice is specified by the useRows and useColumns parameters These two parameters can be set to either even or odd indicating that pixels should be placed in the even or odd rows columns of the input image All other pixels not falling within the decimation lattice are set to zero value No filtering is performed by unDecimate Parameter Information outputWidth Type int Default 0 outputHeight Type int Default 0 useRows Type int Default 0 Values Even 0 Odd 1 useColumns Type int Default 0 Values Even 0 Odd 1 borderWidth Type int Default 0 C Prototype CorOpRtn cor_unDecimate CorImage In CorImage Out int outputWidth int outputHeight int useRows int useColumns int borderWidth Headers
491. whether they are under or over the range The remaining modes provide a linear mapping from an input pixel value range to an output pixel value range All mapping modes work in the same way for float type input images Float type to integer type casts map from the range of pixel values in the input image to the full range of the output image type Float type to float type casts map from the range of pixel values in the input image to the range specified by the min and max parameters In the MapRange mode the full range of the input image type is used as the input image range so integer type to integer type casts map from the full range of the input image type to the full range of the output image type and integer type to float type casts map from the full range of the input image type to the range specified by the min and max parametes In the MapValue mode the range of pixel values in the input image is used as the input image range so integer type to integer type casts map from the range of pixel values in the input image to the full range of the output image type and integer type to float type casts map from the range of pixel values in the input image to the range specified by the min and max parametes The MapMSB mode is similar to the MapValue mode except for integer output images the values are mapped to a range that is at least half the full range of the output image type This mode may 41 have a speed advantage over
492. with direction using ROI Description Hough transform with direction using ROI Parameter Information tolerance Type int Default 5 startAngle Type int Default 128 endAngle Type int Default 127 angleIncr Type float Default 1 rRange Type int Default 0 rResolution Type int Default 0 Values 1 0 2 1 4 2 8 3 16 4 32 5 degrees Type int Default 0 Values 360 0 256 1 Demo iGraph Operators Geometry convexHullL igr C Prototype CorOpRtn cor_houghDirRoi CorImage Magnitude CorImage Orientation CorObj Roi CorImage Out int tolerance int startAngle int endAngle float angleIncr int rRange int rResolution int degrees Headers include WITHOME h wFit h Libraries WITHOME lib wObj lib WITHOME lib wFit lib linelntersect Find the intersection of two lines Description The lineIntersect operator finds the intersection point of two lines or line segments The LineO and Linel inputs must be line type geometric objects Geometry Graphic Edge or Line The extend parameter specifies whether the the line inputs are considered to be finite line segments or infinite length lines extending through the endpoints The Intersect output is a graphic point that indicates the intersection of the input lines if they intersect The Status output indicates whether the lines actually intersect The status value is 0 if the lines do not intersect 1 if they do In a
493. y to time sections of an igraph accurately Parameter Information format Type int Default 0 Values milli 0 micro 1 C Prototype CorOpRtn cor_CPUTime int Time int format Headers include WITHOME h wSystem h Libraries WITHOME lib wObj lib WITHOME lib wSystem lib chdir Change current object directory ay Description The chdir operator allows an executing igraph to change the object directory This way you can use relative path names for operators which access the file system such as wrObj and wrImage Parameter Information filename Type CorFile C Prototype CorOpRtn cor_chdir_1 CorFile Out CorFile filename Headers include WITHOME h wSystem h Libraries WITHOME lib wObj lib WITHOME lib wSystem lib directory Read a directory r ifthe F Description On receiving any object at its input the directory operator forms a vector of strings representing the filenames that match a wildcard filename specifier in the current object directory The filename specification is supplied in the property window and can consist of any valid regular expression eg 205 wit If no matching filenames are found the operator outputs an empty vector This operator is used as a method of sequencing a number of related objects from disk through an igraph see also sequencer Parameter Information filename Type CorFile Default C Prototype CorOpRtn
494. ype Type int Default 0 Values Display 0 Acknowledge 1 Conditional 2 Message Type CorString Default Hello Placement Type int Default 0 Values bitwise OR Fix position 1 X Type int Default 0 Y Type int Default 0 C Prototype CorOpRtn cor_prompt CorObj In int Out int Type CorString Message int Placement int X int Y Headers include WITHOME h wDisplay h Libraries WITHOME lib wObj lib WITHOME lib wDisplay lib runTimeConst Prompt the user for a constant al kb Description The runTimeConst operator is used when the user needs to be prompted for a value during igraph execution The input object triggers this operator to create a pop up window which requires the user to enter a value see constant operator for allowable types When the OK button is clicked an output object is produced with the value Parameter Information label Type CorString Default Enter value Placement Type int Default 0 Values bitwise OR Fix position 1 X Type int Default 0 Y Type int Default 0 C Prototype CorOpRtn cor_runTimeConst CorObj Constant CorString label int Placement int X int Y Headers include WITHOME h wDisplay h Libraries WITHOME lib wObj lib WITHOME lib wDisplay lib 65 smallDisplay Display object data in a pop up window Description smallDisplay performs the same function as display smallDisplay has a sma

OC-WITM-OPREF - Teledyne DALSA Inc

Contents

Download Pdf Manuals

Related Search

Related Contents