Home

AdeptVision Reference Guide, Rev A

1. FLANGE CRANK kk XK BRACKET Related Keywords VSHOW MODEL program instruction VTRAIN program instruction VTRAIN MODEL program instruction VLOAD program instruction VSTORE monitor command VSTORE program instruction VTRAIN FINDER program instruction AdeptVision Reference Guide Rev A 157 VLOAD Syntax Program Instruction VLOAD lun file_spec Function Load vision models ObjectFinder models prototypes Optical Character Recog nition fonts or correlation templates from a disk file Usage Considerations VLOAD will not load files created prior to version 11 0 Parameters lun Sfile_spec Details Real valued expression that specifies the logical unit number to be associated with the operation This must be one of the logical unit numbers for a disk device see the ATTACH instruction in the V Language Reference Guide The logical unit number used must not already be in use by the program for another disk access String expression that specifies the disk file from which the vision models are to be loaded This consists of an optional physical device an optional disk unit an optional directory path a file name and an optional file extension Uppercase or lowercase letters can be used The current default device unit and directory path are considered as appropriate see the DEFAULT command in the V Operating System Reference Guide If no filename extension is specified
2. 168 Vision Coordinate Systems 310 AdeptVision Reference Guide Rev A 5 Table 1 1 Table 1 2 Table 2 1 Table 2 2 Table 2 3 Table 2 4 Table 2 5 List of lables Related Publications 2 2 050 8 AdeptVision Keyword Summary 9 VFEATURE Function Data 96 VFEATURE Function Data for Objectfnder following VLOCATE 102 VFEATURE Function Data for Objecifinder following VSHOW 103 Contents of VGETPIC VPUTPIC Header Sting 143 Elements of VGETCAL VPUTCAL Scaler Calibration Anay 204 AdeptVision Reference Guide Rev A Intoduction Compatibilty 0 0 0 00 2 Conmventons 6 6d aa a a Related Publications 2 2 2 0 0 0 0 0 AdeptVision Keyword Summary 2 eee es How Can IGetHelp 2 ww Within the Continental United States Service Calls 2 2 0 Application Questions Applications Intemet E Mail Address Training Infomation Within Europe o PONCE 65 oP RES SEO A Outside Continental United StatesorEurope AdeptFaxonDemand Adepton Demand Web Page AdeptVision Reference Guide Rev A Chapter 1 Compatibility Compatibility This reference guide is for use with Adept controllers equipped with the A
3. Image of fork split into two regions V SHOW BOUNDS enabled Recognition of fork with V DISJOINT and V SHOW RECOG enabled Figure 2 5 Effect of V DISJOINT Switch Related Keyword V BOUNDARIES system switch V HOLES system switch V TOUCHING system switch 78 AdeptVision Reference Guide Rev A Monitor Command and Program Instruction VDISPLAY Syntax VDISPLAY camera mode overlay xpan ypan zoom Function Select the current vision display mode or the display mode to be used when the vision system performs its normal image processing functions Usage Considerations The VDISPLAY monitor command may be used when a program is executing Vision model training must not be active for this instruction to execute Parameter camera Optional real valued expression that specifies the virtual camera number The special camera number 0 specifies all virtual cameras This parameter defaults to 0 all virtual cameras if no value is specified In that case the parentheses can be omitted mode Real valued expression that specifies the mode of operation to be used The available modes are listed here but described below 1 Grayscale camera output live grayscale 0 Thresholded camera output live binary 1 Frame grabbed grayscale 2 Frame grabbed edge or binary 3 Special display graphics only 4 Static graphics graphics only 5 No display no alter overlay Optional real valued expression that specifi
4. 0 Run state 0 Idle 1 Running 2 Training 3 Frame held 4 Busy 5 Waiting 1 Number of objects recognized 2 Number of regions not recognized 240 Adept Vision Reference Guide Rev A Program Instruction VSTATUS The above set of information is the same as that displayed by the VSTATUS monitor command See the description of the VSTATUS monitor command for an explanation of the values Relative Contents of Array Elements Index for type 1 0 Selected frame 1 or 2 1 Virtual camera 1 through 32 2 Display mode 1 0 1 2 3 4 or 5 3 Overlay mode 0 1 or 2 4 Displayed frame 1 or 2 5 Displayed virtual camera 1 through 32 6 Valid image data 0 false or 1 true 7 Valid run lengths 0 false or 1 true The above items are returned when the VSTATUS type parameter is 1 The first two items Selected frame and Virtual camera are the frame store and virtual camera currently selected The VPICTURE and VSELECT instructions can change the current frame and camera as can menu selections made with the mouse The next four items from a type 1 VSTATUS describe what is being displayed in the Vision display window Display mode and Overlay mode have the same meaning as the corresponding parameters to the VDISPLAY operator Displayed frame indicates which of the two frames is being displayed assuming one is being displayed infer from the current display mode
5. An object is located and properly identified when all seven steps are successfully completed At that time the object s identification position 2 D orientation and goodness of appearance measure verified percentage are sent to the main system processor This information is then available to application programs written in the V language NOTE An additional tool called the ObjectFinder provides an alternate method for locating objects in an image Refer to Appendix C of the AdeptVision User s Guide for information on the ObjectFinder tool Connectivity Analysis 302 This is the common technique developed at SRI International for separating closed regions from the background The processing is performed in a single pass across the image The binary image is scanned and converted into run length encodings on a line by line basis This compact image representation is simply a AdeptVision Reference Guide Rev A Appendix B Chain Encode Perimeters list of the column numbers where there are color transitions of black to white or white to black Connectivity analysis groups contiguous horizontal pixels of the same color into run lengths and then groups contiguous vertical run lengths of the same color thereby completely segmenting the image into black and white regions While processing the run length encodings the algorithm also computes the area and perimeter of each region Finally as regions close off they are i
6. Program Instruction VWINDOWI Optional array index that identifies the first element to be defined in data The default is 0 If a multiple dimension array is specified only the right most index is incremented as the values are assigned Integer value specifying the image buffer region for the inspection window Image buffer regions specify both a size and a frame store see the description of VDEF AOD Optional real valued expression indicating the shape of the window 1 rectangular the default 2 circular or ring shaped Real valued expressions specifying the x y position of the center of the window in millimeters Real valued expressions specifying the width dx and height dy of a rectangular window in millimeters Optional real valued expression specifying the orientation of the window in degrees The default is 0 degrees Real valued expression specifying the outer radius of a circular or ring shaped window in millimeters Optional real valued expression specifying the inner radius of a ring shaped window in millimeters The default is 0 meaning that the window is not ring shaped that is it has no inner circle Optional real valued expressions specifying the angular range of a circular or ring shaped window in degrees Each parameter defaults to 0 A full circle or ring is indicated when both values are 0 A VPICTURE instruction must be issued before window processing so that an image is avail
7. VSTATUS Camera Run number state 1 Idle 2 Waiting 4 Running 5 Idle Related Keyword Objects recognized 79 0 0 0 VSTATUS program instruction Regions not recognized 1 0 0 56 AdeptVision Reference Guide Rev A 239 VSTATUS Program Instruction Syntax VSTATUS camera type array index Function Return vision system status information in a real array Parameters camera Optional real valued expression that specifies the virtual camera number The default is 1 type Optional real valued expression that specifies the group of information wanted The default is 0 See below for details array Real array that receives the data The data are stored in sequential array elements index Optional integer value that identifies the first array element to be defined in array Zero is assumed if the index is omitted If a multiple dimension array is specified only the right most index is incremented as the values are assigned Details This is functionally equivalent to the VSTATUS monitor command except that more information is available and it is placed into an array instead of being displayed There are two different sets of information that can be requested with the type parameter The values stored for each set are as follows the actual element containing a particular item is index plus the item s relative index Relative Contents of Array Elements Index for type 0
8. below Real valued expression specifying the radius range for the search in millimeters This is the distance across the estimated radius within which the vision system searches for the arc s edge points In other words the search extends from r rr 2 to r rr 2 The range of acceptable values for rr is between 0 and 2 r Optional real valued expressions specifying the angular range for the search in degrees If ang0 and angn are 0 and 360 or 0 and 0 the search space is a complete ring and a circle is fit The acceptable range of values for these parameters is 1000 to 1000 Adept Vision Reference Guide Rev A 107 VFIND ARC Program Instruction degrees If angn ang0 exceeds 360 degrees a complete ring is assumed The circumference of the estimated arc must be at least three pixels The arc circumference with r in pixels is 2 n r angn ang0 360 The defaults are 0 Details 108 VFIND ARC operating in grayscale mode finds a circular arc edge with subpixel accuracy The instruction has three basic modes of operation find the center and radius of the arc s circle find only the center knowing the radius dim3 of the area of interest definition or find only the radius knowing the center dim1 and dim2 of the area of interest definition The operator takes about the same amount of processing time for each of the three modes but the accuracy of the results is diff
9. Adept Technology maintains a Customer Service Center at its headquarters in San Jose CA The phone numbers are Service Calls 800 232 3378 24 hours a day 7 days a week 408 433 9462 FAX NOTE When calling with a controller related question please have the serial number of the controller If your system includes an Adept robot also have the serial number of the robot The serial numbers can be determined by using the ID command see the V Operating System User s Guide 14 AdeptVision Reference Guide Rev A Chapter 1 How Can I Get Help Application Questions If you have an application question you can contact the Adept Applications Engineering Support Center for your region Phone Adept Office Hours Region San Jose CA Voice 408 434 5033 Western Region States Fax 408 434 6248 AR AZ CA CO ID KS LA MO MT NE 8 00 A M 5 00 P M PST NM NV OK OR TX UT WA WY Cincinnati OH Voice 513 792 0266 Midwestern Region States Fax 513 792 0274 AL IA IL IN KY MI MN MS ND West 8 00 A M 5 00 P M EST NY OH West PA SD TN WI Southbury CT Voice 203 264 0564 Eastern Region States Fax 203 264 5114 CT DE FL GA MD ME NC NH MA 8 00 A M 5 00 P M EST NJ East NY East PA RI SC VA VT WV Applications Internet E Mail Address If you have access to the Internet you can send application questions by e mail to adeptinfo infolab com
10. Like VRULER and VWWINDOWL VWINDOW instructions queue on the vision processor Thus an application program does not have to wait for picture processing to complete before executing a VWINDOW instruction The values of the parameters V FIRST COL V FIRST LINE V LAST COL and V LAST LINE are used to clip the window to the image That is the vision system ignores any portion of the window that is outside these boundaries Type 0 rulers and finders and clear grip tests are clipped to the bounds defined by the latest VWWINDOW instruction Only transitions within those bounds are found Rulers and finders other than type 0 and inspection VWINDOWI windows are not limited by the bounds defined by the latest VWINDOW Example Perform a quick frame grab and process two windows The first window is centered at the point 140 260 20mm x 30mm and rotated 45 degrees The second one is centered at 300 330 40mm x 40mm with no rotation The virtual camera for the first window is 2 and the virtual camera for the second window defaults to 1 VDEF AOI 2000 1 140 260 20 30 45 VDEF AOI 3000 1 300 330 40 40 VPICTURE cam 2 VWINDOW 2 2000 VWINDOW 3000 Related Keywords VDEEAOI program instruction VPICTURE monitor command and program instruction VWINDOWB program instruction VWINDOWI program instruction AdeptVision Reference Guide Rev A Program Instruction VWINDOWB Syntax VWINDOWB cam m
11. Para meter index Real value specifying the desired information item The value must be in the range 1 to 50 inclusive see below Details The VFEATURE indexes corresponding values and units of measure are listed in Table 2 1 Table 2 2 and Table 2 3 All are explained in detail below Table 2 1 VFEATURE Function Data Index Value Units comment 1 Valid TRUE FALSE 2 X millimeters 3 Y millimeters 4 Z millimeters 5 RX degrees 6 RY degrees 7 RZ degrees 8 Encoder offset encoder counts 9 Verify percentage percentage 10 Area raw pixels 11 Region ID 12 Instance ID 96 AdeptVision Reference Guide Rev A Real Valued Function VFEATURE Table 2 1 VFEATURE Function Data Continued Index Value Units comment 13 Min_X millimeters 14 Max_X millimeters 15 Min_Y millimeters 16 Max_Y millimeters 17 holes in region 18 Time seconds 19 Flags bit field TRUE FALSE 20 Valid gripper number of the gripper 21 Hole number number of the hole 22 Parent number number of hole s parent 23 Virtual camera number 1 through 32 24 Effort level 0 through 4 25 Color of prototype 0 or 1 26 samples taught 27 bounds 28 Min area of prototype raw pixels 29 Max area of prototype raw pixels 30 Cameras of prototype bit field 1 to 16 31 Cameras of prototy
12. VFEATURE Return specified information about the object most recently VLOCATEd or the prototype most recently displayed by the VSHOW program instruction VFIND ARC Fit a circular arc to an image edge bounded by a window that is shaped like a ring or a ring segment VFINDER Perform ObjectFinder recognition using the planning associated with the given virtual camera 10 Adept Vision Reference Guide Rev A Chapter 1 AdeptVision Keyword Summary Table 1 2 AdeptVision Keyword Summary Continued Keyword Description VFIND LINE Fit a straight line to an image edge within a window VFIND POINT In a search window find an edge point that is nearest to one side of the window V FIRST COL Set the number of the first column of frame store data to be processed V FIRST LINE Set the number of the first line of frame store data to be processed V FIT ARCS Enable or disable the fitting of circular arcs when performing boundary analysis V GAIN Set the gain for the incoming video camera signal VGAPS Find the unverified gaps in a match with a prototype or subprototype VGET AOI Return the definition of an area of interest VGETCAL Ask the system to fill in arrays with the previously defined vision calibration data for a given virtual camera VGETPIC Read all or part of an image into a string array VGET TRANS Return the value of the current vision transformation V
13. VPICTURE cam 2 VDEF AOI 2000 1 cx Cy dx dy ang VOCR cam op dm data locs font Sexp 2001 al AdeptVision Reference Guide Rev A 187 VOCR Program Instruction TYPE chars found data 0 S TYPE Average score data l Min score data 2 TYPE chars found data 0 S TYPE Average score datal l1 Min score data 2 IF op 1 OR op 2 THEN TYPE Expected 1st picks data 3 S TYPE lst or 2nd data 4 END IF op gt 2 THEN If individual picks returned cc 5 FOR cnt 1 TO data 0 TYPE Picks SINTB data cc S TYPE 10 data cct1 S TYPE SINTB data cc 2 S TYPE 10 data cc 3 T cc cct4 END END Draw text bounding box GTRANS 20 1 GRECTANGLE 20 locs 0 locs 1 locs 2 locs 3 IF op gt 2 THEN FOR cnt 0 TO data 0 1 Dot centers of regions GARC 20 locs 2 cnt 4 locs 2 cnt 5 3 END END PROMPT Press Enter to repeat or Ctrl Z to quit Sans END END Related Keywords 188 VDEF AOI program instruction VDEEFONT program instruction V MAX AREA system switch V MAX TIME system switch V MIN AREA system switch V MIN HOLE AREA system switch V SHOW EDGES system switch VSHOW MODEL program instruction VTRAIN MODEL program instruction AdeptVision Reference Guide Rev A System Parameter V OFFSET Syntax V OFFSET camera
14. locs 1 is specified it receives a definition of the Adept Vision Reference Guide Rev A 185 VOCR Program Instruction 186 minimum box that surrounds all of the bounded areas If the operation is Robust Verification Full Output op 2 or Recognition op 3 the center of the bounding box for each character is also returned All of the position information is in millimeters Information returned by all operations loc j 0 Minimum X of text bounding box loc j 1 Minimum Y of text bounding box loc j 2 Maximum X of text bounding box loc j 3 Maximum Y of text bounding box Additional information returned when op is 2 or 3 loc j 4 X component of center of character bounding box loc j 5 Y component of center of character bounding box These two items repeat for each character region found and analyzed a total of data i 0 times Any bounded area taller than 63 pixels is ignored If the characters to be recognized are larger than this normal prototype recognition may be used VOCR also considers the values of V MIN AREA V MIN HOLE AREA and V MAX AREA These parameters may be used to filter unwanted regions within the VOCR tool search area Up to 80 character regions are analyzed in an OCR window Any additional bounded areas are ignored Similarly when verifying text extra characters in the window are ignored For example if there are 10 characters in the given Sexpected st
15. parameter as described below AdeptVision Reference Guide Rev A 287 VWINDOWB Program Instruction index Optional array index that identifies the first element to be defined in datal The default is 0 If a multiple dimension array is specified only the right most index is incremented as the values are assigned ibr Optional integer value specifying the image buffer region for the inspection window Image buffer regions specify both a size and a frame store see the description of VDEF AOI A VWINDOWB tool must be an orthogonal rectangle so the image buffer region s AOI must use a rectangular shape and any angle specifications are ignored Default is current frame full size image buffer region 1000 Details 288 This instruction quickly calculates basic window statistics Since all regions are considered as one blob execution time is independent of scene complexity When calculating perimeter VWWINDOWB counts any foreground pixel as a perimeter pixel if it has one or more background pixels as neighbors in a 3x3 neighborhood A VPICTURE instruction must be issued before window processing so that an image is available for analysis Any VPICTURE mode may be used except future frame grab mode 1 Quick frame grab mode 2 is the recommended mode The window types are 1 Returns the area bounding box and centroid of a blob 2 Returns the area bounding box centroid and perimeter of a blob perimeter a
16. pix to pmm must also be specified Optional two dimensional real array that defines the pixel to millimeter transformation matrix If this array is specified the array pmm to pix must also be specified Optional integer values that identify the first array element to be referenced in the respective array Zero is assumed for each index that is omitted If an array is specified that has more dimensions than needed only the right most indexes are incremented as the values are referenced Optional transformation expression that defines the vision transformation to store with the calibration data Note in Vt version 11 0 to cam must be a variable In version 11 1 and higher to cam will not have to be a variable The calibration programs supplied with the AdeptVision systems fill the calibration arrays and execute a VPUTCAL instruction at the end of a successful calibration This calibration data is used when performing operations that involve millimeter distances or locations in the image such as VPICTURE VWINDOW VRULER VLOCATE and VFEATURE Since the calibration arrays are generally filled by an Adept utility program the programmer should not have to be familiar with all the individual array elements However since some elements may be useful to know about a brief description of the array elements is provided in the table below Adept Vision Reference Guide Rev A 203 VPUTCAL Program Instruction Any undef
17. prototype cx cy dx dy ang value Initiate training of the prototype whose name is specified cam mode arg prototype ibr shape VTRAIN FINDER cam mode dmode arg arg2 arg3 model_name ibr value value2 value3 Initiate training of the finder model whose name is specified VTRAIN MODEL Stext ibr VTRAIN MODEL cam plan model_name Stext shape cx cy dx dy ang Train on a vision model a correlation template or an Optical Character Recognition OCR font For correlation this instruction defines the template For OCR this instruction trains the vision system to recognize characters in a font or causes the vision system to plan the recognition strategy for a fully trained font cam plan model_name VWAIT type ibr Delay program execution until processing of a VPICTURE or VWINDOW operation is complete Adept Vision Reference Guide Rev A Appendix A AdeptVision Quick Reference VWINDOW cam type dmode how_many ibr VWINDOW cam type dmode how_many shape CX Cy width height ang VWINDOWB cam mode dmode data index ibr Extract image information from within a rotatable rectangular window VWINDOWI cam type dmode sample da ta index ibr VWINDOWI cam type dmode sample da talindex shape cx cy dx dy ang VWINDOWI cam type dmode sample da ta index shape cx cy or ir ang0 angn Extract image information from within a windo
18. Finally the Displayed virtual camera is the virtual camera associated with the displayed frame The last two items from a type 1 VSTATUS indicate whether the selected frame contains valid grayscale image data or valid run lengths Valid run lengths is true if a VPICTURE in mode 1 or 0 or a VWINDOW operation was performed If run lengths are valid a type 0 rulers and finders instruction may be performed The camera parameter is not relevant for VSTATUS type 1 VSTATUS is a nonsequential instruction The distinction between sequential and nonsequential processing applies only when multiple V tasks are executing vision instructions Then while the vision system is processing a vision instruction for one task and additional vision instructions are queued up from other tasks another task can execute a VSTATUS instruction and it will complete immediately before completion of the instruction that was being processed or of queued instructions Adept Vision Reference Guide Rev A 241 VSTATUS Program Instruction Example Store type 0 status information for virtual camera 1 in the array status in elements 10 through 12 VSTATUS 1 status 10 Related Keyword VSTATUS monitor command 242 AdeptVision Reference Guide Rev A Syntax Monitor Command VSTORE VSTORE file spec model_name model_name Function Store in a disk file selected or all vision prototypes and their subproto
19. Function Set the offset for the incoming video signal that is program the zero reference for the A D converter Usage Considerations Changing this parameter immediately affects the video output of the camera interface board This is an array of switches one for each virtual camera See the AdeptVision User s Guide for details on setting switches Details The V OFFSET parameter works with the V GAIN parameter to select the incoming analog video offset and gain respectively V GAIN multiplies scales the video signal whereas V OFFSET shifts translates the video signal Before adjusting the values of VOFFSET and V GAIN you should take a picture compute the histogram and study the video data displayed in the Vision display window To take the picture and compute the histogram enter the V monitor commands VPICTURE cam 2 and VHISTOGRAM where cam is the number of the virtual camera being used Or use the mouse to make the menu selections to perform these same operations The goal is to have the video data fill most of the intensity range 0 to 127 without spilling over either end If the video data spills over the left end the histogram curve shows a spike over the 0 intensity label Similarly if the video data spills over the right end the curve shows a spike at or near the 127 intensity label You should increase V GAIN to expand the intensity range of the video data or decrease V
20. OCR font font_num chars height VDEFGRIP proto grip mode num_fngrs trans i w 3 h x VDEFGRIP proto 0 Define the shape and position of a robot gripper for clear grip tests VDEF LUT lut num lut Define a grayscale and binary look up table for mapping graylevel and binary values during a VCOPY operation VDEF MORPH mode type array dx dy Define a binary morphological operation VDEF SUBPROTO proto subname first_edge last_edge Define a subprototype VDEF TRANS mode dx dy angle scale Define a transformation to apply to the location of all vision tools placed until the next VDEF TRANS instruction VDELETE model_name Delete a specified prototype subprototype Optical Character Recognition OCR font or correlation template in the vision system V DISJOINT camera Determine whether or not prototypes may be matched to multiple disjoint regions 295 VDISPLAY VDISPLAY camera ypan zoom Select the current vision display mode or the display mode to be used when the vision system performs its normal image processing functions mode overlay xpan V DRY RUN Enable graphics only mode for various vision operators VEDGE cam type dmode dest_ibr src_ibr Compute edges in the grayscale image and threshold the edges replacing the binary image using either a cross gradient or Sobel algorithm VEDGE INFO data i proto_nam edge_num Retrieve information about the edge
21. The results of rulers line and arc and finders line point and arc are transformed to millimeters using the pix to pmm transformation The results of all other vision tools are transformed using the simpler scale transformation If correction for perspective distortion is disabled the pmm to pix and pix to pmm transformations that VGETCAL returns represent the following simple scale transformations 1 X scale 0 pmm to pix 0 1 Y scale 0 0 X scale 0 pix to pmm 0 Y scale 0 0 Both of the arrays pmm to mm and mm to pmm contain identity transformations when correction for perspective distortion is disabled How to Use the Anays pmm to mn and mm to pmm How do you use pmm to mm and mm to pmm to correct for perspective distortion when the vision system does not automatically do it for you The V routine tr point see below transforms a point using the transformation array provided Your application program can use this routine as follows e Lines are transformed by simply transforming the two endpoints of the line e Circles transform to an egg shape under perspective projection NOTE The effect can be approximated by transforming the center of the circle A radius for the transformed circle can then be computed by transforming some points on the perimeter of the circle and c
22. active or has busy device drivers If the system has no robot only a single execution task running and no busy device drivers the delay can be as long as 2 milliseconds Adept Vision Reference Guide Rev A 149 V IO WAIT System Parameter What is the effect of a worst case delay of 16 milliseconds Consider for example a field of view that is 5 12 centimeters wide Each pixel then is 0 1 millimeter wide If the objects in the field of view are moving at the high speed of 1 meter per second a worst case delay of 16 milliseconds corresponds to a shift of 160 pixels Since the frame store is 512 pixels wide a shift of 160 pixels could shift part of the object being analyzed out of the field of view When the V IO WAIT parameter is set to 1 a VPICTURE operation could wait indefinitely for the external trigger However the VPICTURE instruction has a special no wait mode that may be used to allow the application program to continue without waiting for the VPICTURE operation to begin Also note that the VPICTURE monitor command or a program containing the VPICTURE instruction may be aborted in various ways When V IO WAIT is set to 0 the default VPICTURE operations do not wait for the external trigger This parameter may be assigned only the values 0 or 1 Example Wait for the external trigger when taking pictures with camera 1 PARAMETER V IO WAIT 1 1 Related Keywords VPICTURE monitor command and progr
23. and AdeptVision systems are loaded into memory from disk Example Start processing camera images beginning with the first line of data PARAMETER V FIRST LINE 1 Related Keywords VDEEAOI program instruction V FIRST COL system parameter V LAST COL system parameter V LAST LINE system parameter 130 AdeptVision Reference Guide Rev A System Switch V FIT ARCS Syntax V FIT ARCS camera Function Enable or disable the fitting of circular arcs when performing boundary analysis Usage Considerations A change to this switch takes effect when the next VPICTURE VWINDOW or VTRAIN operation is executed This is an array of switches one for each virtual camera See the general description of switches in the AdeptVision User s Guide for syntax details Details This switch allows the user to disable arc fitting during boundary analysis When arc fitting is disabled the vision system attempts to fit only lines instead of attempting to fit both lines and arcs This improves performance for scenes that contain no or very few boundary segments that appear as arcs If the vision system is doing prototype recognition and none of the prototypes associated with the virtual camera being VPICTUREd have any arcs V FIT ARCS is automatically disabled for the VPICTURE or VWINDOW operation Even if the prototypes have arcs the vision system will fit only arcs with similar radii filtering out arcs with other radi
24. disk takes 13 5 seconds Copying it to a frame buffer takes about 0 1 seconds If your controller is equipped with the AdeptVision Enhanced VXL Interface option binary correlation may be performed using the hardware binary correlator This executes a very fast correlation operation The area of interest for binary correlation must be at least 16 pixels wider and 4 pixels higher than the template With 640x480 virtual frame stores unused portions of the frame store are used to store the correlation template This unused space is 384x480 pixels so this is the largest template allowed For a hierarchical search the template is limited to 384x341 pixels For frame stores smaller than 640x480 the template may be as large as the virtual frame store although hierarchical correlation is limited to 512x341 pixels Hierarchical correlation and binary correlation require an extra scratch frame store If both physical frame stores contain image data VSELECT mode 1 must be used to specify which frame store to use as the scratch frame store 1 An effective hierarchical search has a depth that quickly identifies candidate areas Too shallow a search will be slow to identify candidate areas The give_up parameter can also help the effectiveness of a search by halting unproductive searches before they reach the top level 48 AdeptVision Reference Guide Rev A Program Instruction VCORRELATE Example Performs a normalized cor
25. except for the base orientation is also returned in the datal parameter of this instruction Examples Display a list of all defined templates VSHOW MO DEL TMPL_0 Get all the information about font font num but don t display it VSHOW MO DEL Related Keywords 236 1 Sstr data FONT_ SENCO VCORRELATE program instruction VDEEFONT program instruction VOCR program instruction VTRAIN MODEL program instruction AdeptVision Reference Guide Rev A DE I0 font num System Switch V SHOW RECOG Syntax V SHOW RECOG camera Function Enable the special display of the objects recognized Usage Considerations A change to this switch takes effect when the next region is displayed in the Vision display window with a graphical display mode in effect This is an array of switches one for each virtual camera See the AdeptVision User s Guide for details on setting switches Details This switch is normally enabled so that the operator can see what the vision system is recognizing That is when this switch is enabled the silhouettes of all recognized prototypes are drawn in the Vision display window overlaying the displayed regions The silhouettes are displayed in a unique color to distinguish them from other graphics If no prototypes are loaded in the vision system the switch has no effect Exa mple Show the objects recognized ENABLE V SHOW
26. locs j font_num Sexpected ibr VOCR cam op dmode data i locs j font_num Sexpected shape cx cy dx dy ang Function Perform Optical Character Recognition OCR or text verification in a rectangular image window Usage Considerations A font must have been created previously with the VDEF FONT and VTRAIN MODEL instructions Parameters cam Optional real valued expression indicating the virtual camera number to use for selecting various system parameters such as V MIN AREA and V MAX AREA The default camera is 1 op Optional real valued expression specifying the desired operation to perform the default is 0 0 Quick verification of expected text 1 Robust verification of expected text short output 2 Robust verification of expected text full output 3 Recognition no expected text dmode Optional real valued expression specifying the display mode to use when displaying the border of the area of interest 1 no draw 0 erase 1 draw solid 2 complement 3 draw dashed 4 complement dashed The default is 1 draw solid NOTE The parentheses in the instruction syntax can be omitted if all three of the above parameters are omitted data Real array that is filled with data describing the results of the OCR operation see below locs Optional real array that is filled with data describing the position of the recognized characters see below ip 3 Optional array index that identifies t
27. mask value 1 Don t display constraints 0 Location constraints are not displayed Display constraints 1 Any position or orientation constraints associated with the prototype are shown in the Vision display window Mode bit 4 must be 0 so that the prototype is displayed After the prototype is VSHOWed the prototype s nominal location and relative constraints can be seen overlaying live video use the monitor command VDISPLAY camera 1 1 to see how the prototype s nominal location compares with the object currently in the camera field of view Bit 2 Get next 0 versus Get specific 1 mask value 2 Get next 0 If bit 3 is set get hole mode see below the next hole is returned If bit 3 is clear and bit 6 is set get subproto mode the next subprototype is returned If bits 3 and 6 are both clear the next prototype is returned If the next prototype or subprototype is requested the proto_name parameter must be a string variable It is filled in with the name of the prototype or subprototype found A VSHOW in get first mode see bit 5 below must be executed before the next prototype can be found If the last prototype hole or subprototype was already returned VFEATURE 1 returns FALSE after a get next VSHOW Get specific 1 A specific prototype or subprototype is accessed Its name must be specified by proto_name as a string expression Bits 3 5 and 6 are ignored in this
28. the extension VS is appended to the name given All the vision models previously VSTOREd in the given file are loaded and added to those already in memory If the file contains a vision model with the same name font number or template number as one already in the vision system loading is aborted That is none of the vision models in the file are loaded Any subprototypes associated with a prototype are also restored automatically Vision models already in RAM cannot be overwritten by new models with the same name They must be first be VDELETEd or VRENAMEd 158 AdeptVision Reference Guide Rev A Program Instruction VLOAD The IOSTAT real valued function can be used after this instruction to determine if any error occurred during the load operation see the V Language User s Guide for details NOTE The application program must not have attached the logical unit since the VLOAD instruction automatically attaches and detaches the logical unit Example The following program instruction loads the models stored in the disk file named OBJECTS VS on the default system disk Logical unit number 6 is associated with the operation and is used to check for successful completion VLOAD 6 objects F IOSTAT 6 lt 0 THEN TYPE C1 VLOAD failure SERROR IOSTAT 6 C1 HALT END Related Keywords VLOAD monitor command VSTORE monitor command VSTORE program instruction Adep
29. the vision system ignores any portion of the window that is outside these boundaries If V RECOGNITION is enabled prototype recognition is being performed the how_many parameter may be very effective in reducing processing time in VPICTURE modes 4 1 and 0 The how_many parameter is not applicable in other modes If the image contains extraneous edges or is noisy the system normally locates all instances of prototypes quickly but then spends a relatively long time analyzing the extraneous edges The extra time analyzing the extraneous edges can be avoided if the programmer knows that there is a certain number of objects in the image and specifies that number as the how_many parameter AdeptVision Reference Guide Rev A Monitor Command and Program Instruction VPICTURE The V TOUCHING and V DISJOINT switches affect the way the how_many parameter is interpreted V TOUCHING affects the way how_many applies to each region in the image whereas V DISJOINT affects the way how_many applies to the image as a whole If the V TOUCHING switch is enabled up to how_many objects are recognized per region If V TOUCHING is disabled no more than one object is recognized per region in the image If V DISJOINT is enabled at most how_many objects are recognized in the entire image If V DISJOINT is disabled there is no limit to how many objects are recognized in the entire image The following table summar
30. the vision system knows what text to expect With text recognition the vision system tries to recognize the text with no expectations except that the characters in the image can be found in the designated font Before this instruction is used the specified OCR font must have been defined using the VDEF FONT instruction and fully trained using the VTRAIN MODEL instruction See the AdeptVision User s Guide for an overview of the AdeptVision OCR capability VOCRK operates on the binary image in the current frame store The virtual camera argument to VOCR selects the group of system switches and parameters to use The following switches and parameters affect VOCR V MIN HOLE AREA V MIN AREA V MAX AREA V MAX TIME and V SHOW EDGES The system switch V SUBTRACT HOLES is always enabled during VOCR operations Adept Vision Reference Guide Rev A 183 VOCR Program Instruction 184 The VOCR image buffer region should encompass a single line of text Within the image buffer region characters are ordered left to right assuming no rotation element 5 of the AOI is 0 If rotated the definition of left changes with the rotation The VOCR operator is drawn like VWINDOW or VWINDOWL except that the left edge is drawn in a special color for reference To see the character outlines during recognition enable V SHOW EDGES for the specified virtual camera and provide a dmode value other than 1 no draw Note however that t
31. with the first object includes the time spent to acquire the image Other consequences of this timekeeping procedure are as follows If multiple objects are recognized in one region the time spent performing boundary analysis is attributed only to the first object located Furthermore if the V DISJOINT switch has been enabled the time spent analyzing the boundaries of all regions is attributed to the first object recognized The elapsed time reported for subsequent objects is only the time spent in the recognition algorithm ITEM 19 Flags is a bit field Bit 1 is reserved for future use Bit 2 is defined when the V HOLES switch has been enabled If set the bit indicates that the region or hole VLOCATEd helped to verify the recognition of the object The boundary of a region helps to verify the recognition if the region boundary coincides with the prototype boundary Only part of the hole boundary must coincide in order to help The other bits in the bit field are reserved for future use ITEM 20 Following a VLOCATE this is the number of the grip position that is clear This VFEATURE entry is zero if no grippers have been defined via VDEFGRIP or if none of the grip positions are clear Otherwise it is the number of the first grip that is clear Grips are tested in the order of their numbering 1 to 4 ITEMS 21 and 22 Hole number and parent number provide a numbering of holes for determining parent child and sib
32. 1 Get information about the prototype edge from VEDGE INFO Then get information about the edge gaps from VGAPS VEDGE INFO einfo proto enum VGAPS gaps enum TYPE C1 Corner einfo 2 einfo 3 IF einfo 0 O THEN TE Line TYPE enum Line gaps 1 verified ELSE Else arc IF einfo 0 gt 0 THEN Stype Convex ELSE Stype Concave END TYPE enum Stype arc S TYPE Radius einfo 8 S TYPE center einfo 6 einfo 7 S TYPE gaps 1 verified d360_circum 360 2 PI einfo 8 AdeptVision Reference Guide Rev A Program Instruction VGAPS cang ATAN2 einfo 3 einfo 7 einfo 2 einfol 6 END For each gap in the edge type out its range and label the gap with a G TYPE Gaps S FOR gnum 1 TO gaps 0 IF gnum gt 1 THEN TYPE S END mid gaps gnum 2 gaps gnum 2 1 2 IF einfo 0 THEN dx einfo 4 einfo 2 dy einfo 5 einfo 3 mid mid SQRT SQR dx SOR dy gcx einfo 2 dx mid gcy einfo 3 dy mid IF einfo 0 gt 0 THEN mid cang mid d360_circum mid cang tmid d360_circum gcx einfo 6 teinfo 8 COS mid gcy einfo 7 einfo 8 SIN mid RANS vwin 1 GTYPE vwin xta 0 xta 1 G DECOMPOSE xta vloc TRANS gcx gcy 0 G TYPE
33. 133 VGAPS Syntax Program Instruction VGAPS data i proto_name edge_num Function Find the unverified gaps in a match with a prototype or subprototype Usage Considerations The V LAST VER DIST system parameter must be nonzero when the last VPICTURE was performed to make the necessary information available to the VGAPS instruction The VGAPS instruction refers to the object most recently VLOCATEd regardless of which program task executed the VLOCATE instruction Consequently for predictable operation only one program task should execute VLOCATE instructions Parameter 134 data proto_nam Real array containing the requested gap information data i 0 Number of gaps in the edge or edge numbers in the list data i 1 Verify percentage 0 to 100 If the parameter edge_num is O see below this is the verify percentage for the entire prototype Otherwise it is the verify percentage for the one edge data i 2 If edge_num is 0 this is the start of a list of the edges of the prototype or subprototype that have gaps in them If edge_num is not zero this is the start of a list of the millimeter ranges for the gaps in the specified edge where data i 2n and data i 2n 1 indicate the start and end of gap n respectively Optional integer value that identifies the first array element to be defined in data Zero is assumed if the index is omitted If a multiple dimensi
34. 2ND THRESH and V THRESHOLD define a range of grayscale values between which camera pixels are considered to be black Pixel values outside this range are considered to be white For example if the value of V 2ND THRESH is 50 and the value of V THRESHOLD is 70 camera pixels in the range 50 to 70 are interpreted as being black and pixels in the ranges 0 to 49 and 71 to 127 are interpreted as being white The V BACKLIGHT system switch determines whether white or black is the background color Normally one threshold is sufficient for binary image processing V 2ND THRESH should then be assigned the value 0 to disable its use Dual thresholds are sometimes very useful however For example suppose the task is to recognize a black and white checkerboard on a middle gray background By setting the V THRESHOLD and V 2ND THRESH values so that the grayscale range includes the middle gray background the checkerboard will appear as one large square in the scene As an another example the object to be recognized may be a uniform middle gray color that is lying on a checkerboard background In general the correct values for V THRESHOLD and V 2ND THRESH depend on the particular application The program instruction VAUTOTHR may be used to automatically determine thresholds This parameter must be assigned an integer value in the range 0 to 127 inclusive The parameter is set to 0 that is its effect is disabled when the V and AdeptVision systems ar
35. Adept 1850 Adept 1850XP A Series S Series Adept MC Adept CC Adept IC Adept OC Adept MV AdeptVision AIM VisionWare AdeptMotion MotionWare PalletWare FlexFeedWare AdeptNet AdeptFTP AdeptNFS AdeptTCP IP AdeptForce AdeptModules AdeptWindows AdeptWindows PC AdeptWindows DDE AdeptWindows Offline Editor and V are trademarks of Adept Technology Inc Any trademarks from other companies used in this publication are the property of those respective companies Printed in the United States of America U D gt N Table of Contents Intoducton 2 Laaa a 7 Compatibilty 6 6464444 bee eee He we ee HR 8 CONVGINIONS 6 4 4 Fore y fash ake Re ee A ee OH HO 8 Related Publications 2 2 2 2 0 0 0 0 0 8 AdeptVision Keyword Summary 2 2 0 a 9 HowCan IGetHelp 2 14 Within the Continental United States 14 Service Calls 5 4 4 4 46 66 6 6469 A 14 Application Questions 2 00 es 15 Applications IntemetE MailAddress 15 Training Information 15 WithinEurope 2 2 2 2 o 15 NONE arrasa AIR 15 Outside Continental United States or Europe 16 Adept Fax on Demand 04 16 Adepton Demand Web Page 16 Desc riptions of Vision Keywords 17 AdeptVision Quick Reference 295 Prototype Recognition Algor thms 301 OV
36. If an object is completely outside the inner border and partially outside the outer border no match proposals are made to that object AdeptVision Reference Guide Rev A 35 V BORDER DIST System Parameter If an object is partially inside the inner border and partially outside the outer border a correct match will probably still be proposed based on features in the reduced image area The match will be verified based on all the visible boundary of the object As a general rule VBORDER DIST should be set to one half the image width of the object being located You may want to experiment with different values of V BORDER DIST To do that you should enable the V SHOW VERIFY system switch select a special display mode VDISPLAY mode 3 and process an image VPICTURE 1 VPICTURE 2 followed by VWINDOW All the attempts to verify match proposals are displayed in the Vision display window Then if the vision system attempts to verify matches to objects cut off by the sides of the image you should increase V BORDER DIST However you should not make V BORDER DIST so big that the system does not recognize objects that are completely visible This parameter must be assigned an integer value in the range 0 to 100 inclusive The value is interpreted as the number of pixels between the outer border and the inner border The parameter is set to 0 that is no border reduction will occur when the V and AdeptVision systems are loaded into memory f
37. Instruction VTRAIN FINDER Syntax VTRAIN FINDER cam mode dmode arg arg2 arg3 Smodel_name ibr value value2 value3 Function Initiate training of the finder model whose name is specified Usage Considerations This instruction requires that all virtual cameras be idle that is no VPICTURE or VWINDOW operation can be executing Training is sensitive to camera calibration Before training begins mount focus and securely fix your cameras and lenses Then calibrate the camera s that will be used to train and recognize the prototypes The same camera s with the same calibration and position used for training must also be used for recognition Parameter cam Optional integer specifying the virtual camera and associated calibration switches and parameters to use when acquiring an image for defining a prototype This camera will also be the default camera associated with the prototype see the AdeptVision User s Guide for details on associating cameras with a prototype mode Optional integer specifying the operating mode for this instruction See below for details dmode Optional real valued expression that specifies the display mode to use when displaying the border of the window 1 No draw O Erase 1 Default Draw solid 2 Complement 3 Draw dashed 4 Complement dashed arg arg2 arg3 Integers whose meaning depends on the value of the mode parameter See below for details Smodel_name String containing the name for the
38. Keep points within 1 pixel Adept Vision Reference Guide Rev A VFIND LINE Program Instruction rect_ibr 2011 VDEF AOI rect_ibr 1 300 270 80 90 VFIND LINE cam 100 a rect_ibr VDEF TRANS 80 70 0 O 90 VF IND 100 b rect_ibr LINE cam p y IF NOT a 0 AND b 0 THEN TYPE Both edges not found STOP El END CALL line_line a 2 X Y is the precision corner point found PROGRAM line _line x1 yl angl x2 y2 ABSTRACT LOCAL dx1 dyl dx2 dy2 f fract dx1 COS dyl SIN dx2 COS dy2 SIN numerator angl angl ang2 ang2 y2 y1 dx2 x2 x1 dy2 aa IF ABS dx1 gt ABS dy1 TH fract dy1 dx1 f numerator fract dx2 dy2 x x1 f y yl fract f El fract dx1 dyl f numerator dx2 fract dy2 y yl f x xl fract f END Related Keywords VDEEAOI program instruction VFIND ARC program instruction VFIND POINT program instruction V MAX PIXEL VAR system parameter V MAX SD system parameter V SHOW EDGES system switch AdeptVision Reference Guide Rev A ang2 X numerator Move the aoi X y y Find the intersection point of two lines N Divide by larger number 123 VFIND POINT Program Instruction Syntax VFIND POINT cam pos dmode effort type data i ibr VFIND POINT cam pos dmode effort type data i 1 xc yc length
39. LINE is 240 AdeptVision Reference Guide Rev A 153 V LAST VER DIST System Parameter Syntax V LAST VER DIST camera Function Enable an extra verification of prototype to image matches and specify the pixel tolerance to use when determining boundary coincidence Usage Considerations A change to this parameter takes effect when the next VPICTURE command or instruction or VWINDOW instruction is executed This is an array of parameters one for each virtual camera See the general description of parameters in the AdeptVision User s Guide for syntax details Details When this parameter is nonzero it enables an extra verification of each prototype to image match Furthermore the value of V LAST VER DIST is used in place of the value of the parameter V MAX VER DIST during the extra verification as the distance criterion in the coincidence test that is when the positions and orientations of prototype edges are compared with those of nearby edges in the image In addition when V LAST VER DIST is nonzero information about the final verification is saved for use by the program instructions VGAPS and VSUBPROTO The VGAPS instruction provides information about unverified gaps in prototype edges When this kind of information is needed the parameter V LAST VER DIST may be assigned a smaller value than VMAX VER DIST thereby specifying a tighter tolerance test However it should not be assigned a value smaller than V MAX
40. LINE or VFIND ARC locates edge points within its area of interest Then the following steps are performed 1 An edge either a line or an arc is fit to the edge points using a least squares technique 2 The distance from each edge point to the fit edge is computed and the standard deviation of all the distances is computed 3 Edge points farther than d from the fit edge are discarded where d is defined as follows d MAX V MAX PIXEL VAR cam V MAX SD cam sd V MAX PIXEL VAR is a system parameter and sd is the standard deviation computed in step 2 above VMAX PIXEL VAR should be set to at least 1 when the finders operate in grayscale mode and to at least 1 5 when the finders operate in binary mode This distance threshold keeps the finders from discarding good edge points when the distance standard deviation sd is small 4 Ifno points were discarded or too many points were discarded processing stops Too many points were discarded means fewer than two edge points remain for fitting a line or fewer than three edge points remain for fitting an arc AdeptVision Reference Guide Rev A System Parameter V MAX SD Otherwise go back to step 1 to refit the edge to the remaining points This filtering capability is useful when salt and pepper noise is present in the image within the area of interest for the finder When the noise is not close to the true edge a single filt
41. ObjectFinder tool Except where noted the usage is the same as for prototypes This command may be used even while a control program is executing Parameters proto_name grip edge_num Details Optional name of a prototype or subprototype to be displayed in the Vision display window If no name is specified the list of all defined prototypes is displayed in the Monitor display window If a subprototype is desired this parameter must have the form namel name2 entered without quotation marks where namel is the name of the prototype containing the subprototype name2 Optional real valued expression that specifies the grip position to be displayed along with the prototype The default grip value is 0 which means no grip positions are shown The value 1 causes all grip positions to be displayed A value in the range 1 to 4 displays the corresponding single grip position Optional real valued expression that indicates 0 Do not show the edge numbers 1 Show all edge numbers gt 0 Show specific edge number The type of model being supported is determined by the name which has been trained as either a prototype or an ObjectFinder model The name is checked against the lists of prototypes first for backward compatibility Therefore you should not use the same name for a prototype model and an ObjectFinder model 222 Adept Vision Reference Guide Rev A Monitor Command VSHOW When a single prototype i
42. RECOGNITION and V EDGE INFO are all disabled the system does not perform arc and line fitting at all since it is not needed see below Thus when V RECOGNITION and V EDGE INFO are disabled the V SHOW BOUNDS switch controls whether or not arc and line fitting is done For applications that do not use recognition disabling VSHOW BOUNDS will decrease processing time because arc and line fitting takes a significant amount of time The three conditions that require the system to perform arc and line fitting are 1 If V RECOGNITION is enabled and prototypes are loaded in the system the system must perform arc and line fitting as a normal step in the recognition process 2 If V EDGE INFO is enabled the system must perform arc and line fitting so that the information is available for VEDGE INFO requests 3 If the VSHOW BOUNDS switch is enabled arcs and lines must be fit so that they can be shown in the Vision display window Related Keywords VEDGE INFO program instruction V RECOGNITION system switch Adept Vision Reference Guide Rev A 229 V SHOW EDGES System Switch Syntax V SHOW EDGES camera Function Enable the special display of edges both the primitive edges that are fit to the boundaries of regions and the edge points that are found by the finders VFIND LINE VFIND ARC and VFIND POINT Usage Considerations A change to this switch takes effect when a VFIND LINE VFIND ARC VFIND POINT VPICTURE in m
43. Syntax V LAST LINE camera Function Set the number of the last line of pixels to be processed Usage Considerations A change to this parameter takes effect when the next VPICTURE or VWWINDOW is executed This is an array of parameters one for each virtual camera See the general description of parameters in the AdeptVision User s Guide for syntax details Details This parameter together with V FIRST LINE is used to set the range of pixel lines that are processed during VPICTURE and VWINDOW operations Lines outside the range specified are ignored by the vision system VDEF AOI is the preferred method for defining a processing border for VPICTURE and VWINDOW The value of the parameter V LAST LINE must be greater than the value of V FIRST LINE and must be less than or equal to 480 The parameter V LAST LINE is set to its maximum value when the V and AdeptVision systems are loaded into memory from disk Example Process all camera images up to the last line of data PARAMETER V LAST LINE 480 Related Keywords VDEEAOI program instruction V FIRST COL system parameter V FIRST LINE system parameter V LAST COL system parameter 1 This is the maximum setting The effective value of V LAST LINE is limited by the virtual frame allocation made with the DEVICE instruction or the CONFIG_C utility In field acquire mode lines are numbered from 1 to 240 Thus in that mode the maximum effective value of V LAST
44. This method also enables you to attach a file such as a portion of V program code to your message NOTE Please attach only information that is formatted as text Training Information For information regarding Adept Training Courses in the USA please call 408 474 3246 or send a FAX message to 408 474 3226 Within Europe Adept Technology maintains a Customer Service Center in Dortmund Germany The phone numbers are 49 231 75 89 40 from within Europe Monday to Friday 8 00 A M to 5 00 P M 49 231 75 89 450 FAX France For customers in France Adept Technology maintains a Customer Service Center in Massy France The phone numbers are Adept Vision Reference Guide Rev A 15 Chapter 1 How Can I Get Help 33 1 69 19 16 16 Monday to Friday 8 30 A M to 5 30 P M CET 33 1 69 32 04 62 FAX Outside Continental United States or Europe For service calls application questions and training information call the Adept Customer Service Center in San Jose California USA 1 408 434 5000 1 408 433 9462 FAX service requests 1 408 434 6248 FAX application questions Adept Fax on Demand Adept maintains a fax back information system for customer use The phone numbers are 800 474 8889 toll free and 503 207 4023 toll call Application utility programs product technical information customer service information and corporate information is available through this automated system There is no charg
45. Usage Considerations The frame store currently selected or the image buffer region specified must contain a valid picture Otherwise an error results Adept recommends the first syntax Para meters cam Optional real valued expression that specifies a virtual camera number The system parameters and switches for this virtual camera are used during some modes of processing see below for details The default is 1 type Optional real valued expression that specifies what image statistics are to be computed and returned in the data array The default type is 0 See the details below dmode Optional real valued expression that specifies the display mode to use when displaying the border of the window 1 no draw 0 erase 1 draw solid 2 complement 3 draw dashed 4 complement dashed The default is 1 draw solid sample Optional real valued expression that specifies the sampling density either 1 or 2 If the value is 1 every pixel in the window is considered If the value is 2 every other pixel on every other line is considered The default is 1 NOTE The parentheses in the instruction syntax can be omitted if all four of the above parameters are omitted data Array into which the image information is to be placed The amount of information depends on the type parameter as described below 290 AdeptVision Reference Guide Rev A index ibr shape cx Cy ang or ir ang0 angn Details
46. VOVERLAPPING system switch is enabled That is the actual setting of V TOUCHING is then ignored This is an array of switches one for each virtual camera See the AdeptVision User s Guide for details on setting switches Details If objects touch in the image their regions merge into a single region In this case V TOUCHING should be enabled so that the vision system will attempt to recognize multiple objects per region Otherwise the system will recognize at most one object per region The V TOUCHING and V DISJOINT switches affect the interpretation of the how_many parameter to the VPICTURE and VWINDOW instructions That parameter specifies the maximum number of objects the vision system should try to recognize V TOUCHING affects how the how_many parameter applies to each region in the image whereas V DISJOINT affects how it applies to the image as a whole If the V TOUCHING switch is enabled up to how_many objects will be recognized per region If V TOUCHING is disabled at most one object will be recognized per region in the image 1 The V TOUCHING switch is automatically considered to be enabled whenever the V OVERLAPPING switch is enabled Adept Vision Reference Guide Rev A 263 V TOUCHING System Switch The following table summarizes the relationship between V TOUCHING V DISJOINT and the how_many parameter to VPICTURE and VWINDOW Number of Objects Number of Objects V
47. a region was found cx VFEATURE vf cx Get centroid Cx Cy cy VFEATURE v cy rmindist VFEATURE vf minr dist Get distance and angle to rmaxdist VFEATURE vf maxr dist closest and farthest pts rminang VFEATURE vf minr ang rmaxang VFEATURE vf maxr ang xx CX COS rminang rmindist Draw line to closest point yy cy SIN rminang rmindist GLINE vlun cx cy XX yy XX CX COS rmaxang rmaxdist Draw line to farthest point yy cy SIN rmaxang rmaxdist GLINE vlun cx cy XX yy VLOCATE nam Locate next region in the image END Related Keywords V 2ND MOMENTS system switch V BOUNDARIES system switch V CENTROID system switch V DISJOINT system switch VFEATURE real valued function V HOLES system switch VLOCATE program instruction V PERIMETER system switch Adept Vision Reference Guide Rev A 179 VMORPH Syntax Program Instruction VMORPH cam type dmode thresh dest_ibr count src_ibr Function Perform a morphological transform on a binary image frame Parameters cam type dmode thresh dest_ibr count src_ibr Details Not currently used Integer value indicating the type of morphological operation to perform Type 1 is erosion 2 is dilation and types 8 through 16 are user definable Types 3 though 7 are undefined reserved for future built in operations Optional real valued expression specifying the display mode to use when displaying the bo
48. area is larger than VMAX AREA the region is merged into the background Any holes of the region also become merged into the background because they are the same color as the background However any holes observed in the holes are kept and treated like separate regions CAUTION If multiple objects touch or overlap their combined area in the image is compared with V MAX AREA Consequently if V MAX AREA is set to be larger than the area of one object but smaller than the combined areas of two objects the image of two touching objects is ignored they disappear Adept Vision Reference Guide Rev A 165 V MAX AREA System Parameter This parameter must be assigned an integer value in the range 1 to 1 048 576 1024 1024 inclusive Furthermore the value must be greater than or equal to the value of the parameter V MIN AREA The parameter V MAX AREA is set to 307 200 when the V and AdeptVision systems are loaded into memory from disk Example Ignore regions that are larger than 100 000 pixels in area PARAMETER V MAX AREA 100000 Related Keywords V MIN AREA system parameter V MIN HOLE AREA system parameter 1 The effective maximum of VMAX AREA is the area of the virtual frame buffer Virtual frame buffer sizes are specified with the DEVICE instruction 166 AdeptVision Reference Guide Rev A System Parameter V MAX PIXEL VAR Syntax V MAX PIXEL VAR camera Function During a VFIND LINE or VFIND ARC operation this pa
49. be seen Related Keywords V RECOGNITION system switch V SHOW BOUNDS system switch V SHOW EDGES system switch 238 AdeptVision Reference Guide Rev A Syntax VSTATUS Function Monitor Command VSTATUS Display vision system status information in the Monitor display window Usage Considerations This command may be used even while a main control program is executing Details This command displays key vision status information in the Monitor display window for all 32 virtual cameras However the status line for a particular camera is not displayed if the run state is Idle and the other entries are 0 This reduces the amount of unimportant text displayed Usually only one or a few cameras have any useful status information The following are explanations of the values that are displayed Camera number Run state Objects recognized Regions not recognized Virtual camera number Indicates the current operational mode of a particular virtual camera Idle Running processing images Busy processing a vision tool Waiting waiting for a fast digital input interrupt signal to acquire an image Frame held for future processing or Training Number of objects matched to a prototype Number of closed regions that were not matched to a prototype Both of the above numerical values are zeroed for a particular virtual camera each time a VPICTURE or VWWINDOW operation is performed for that camera Exa mple
50. case AdeptVision Reference Guide Rev A Bit 3 Bit 4 Bit 5 Bit 6 Program Instruction VSHOW Get prototype 0 versus Get hole 1 mask value 4 NOTE Get hole is not applicable for ObjectFinder models This bit is ignored if bit 2 is set get specific mode Get prototype 0 Requests information about a prototype See bit 2 Get next 0 Get hole 1 VSHOW in this mode requests hole information In order for get hole mode to be used a prototype first has to be VSHOWed The subsequent get hole VSHOWS refer to that prototype Each get hole VSHOW returns one hole of the prototype Note When bit 3 is set bit 6 is ignored In get hole mode no string variable is needed for the proto_name parameter to the VSHOW instruction Display 0 versus Don t display 1 mask value 8 Display 0 The model is displayed in the vision display window The model is positioned at its nominal location the location of the first instance of the model in the field of view Don t display 1 Don t display the model in the Vision display window Get next 0 versus Get first 1 mask value 16 This bit is ignored if bit 2 is set get specific mode Get next 0 Requests information about the next item See bit 2 Get next 0 Get first 1 Return the first model in the system or return the first hole or subprototype associated with the model last VSHOWed The specific action depends on the s
51. edge 153 setting left edge 129 setting right edge 152 setting top edge 130 Image brightness and contrast 132 Information training 15 Input parameters using virtual cameras 115 202 273 Instructions for Adept Utility Programs 8 Internet 15 Introduction 7 L Last column processed in image 152 Last line processed inimage 153 Line Finder tool start position and polarity 119 Line Finder tool sample 120 M Match verify 305 Matrix color coding of 279 convolution 57 discrimination 279 elements of a convolution matrix 60 millimeter to pixel transformation 139 203 pixel to millimeter transformation 139 203 Monitor commands VABORT 24 VAUTOTHR 28 VDEFMORPH 69 VDEFSUBPROTO 71 VDELETE 75 VDISPLAY 79 VLOAD 156 VPLAN FINDER 201 VQUEUE 208 VRENAME 212 VSHOW 222 Index VSTATUS 239 VSTORE 243 VTRAIN 265 N NGC Normalized Grayscale Correlation 47 Normalized grayscale correlation 45 47 O ObjectFinder 172 302 model 156 158 225 243 and get hole 227 and get proto holes and get subproto 227 globally storing only 244 246 name restriction 243 used in planning 103 recognition 10 13 114 115 232 arc and line fitting for 232 for multi instance training 114 tool 222 302 and find hole 161 and VLOCATE parameter restrictions 160 162 subprototype and gap information 161 VFEATURE function data for following VLOCATE 102 following VSHOW 103 OCR fonts 277 Optical Character Recognition OCR 10 11 12 13 1
52. frame buffer 11 arc 5 shape 5 diml 35 dim2 40 5 dim3 12 dim4 8 angl 0 ang2 290 mode 16 4 0 start from outside mask 16 find both mask 4 dark inside mask 0 VDEF AOI arc_ibr arc diml dim2 dim3 dim4 angl ang2 VFIND ARC 1 mode 1 data arc_ibr 110 Adept Vision Reference Guide Rev A Program Instruction VFIND ARC Find both Calculate radius and center mode 1 01 0 0 Start rere Dark inside Looks for Search from arc with a dark inside outer radius within the search range Dot indicates search starts at outer radius Dark inner ring indicates search for arc with dark inside within search range dim4 search range arc with dark inside must occur in this range dim1 dim2 o dim3 dia tool center guideline radius of center of ang search range Figure 2 6 Arc Finder Shape If part of the operator falls off the Vision display window that is if part of the search area must be clipped by the virtual frame buffer defined in the image buffer region that situation is noted by the value of data i 1 The vision system still attempts to fit an arc No arc is fit if fewer than three edge points are found or if the resulting arc radius would be too large Then data i 0 returns FALSE Edge points are found within the search area and then an arc is fit to the edge points using a special triangulation algo
53. in uppercase and the arguments are shown in lowercase The keyword must be entered exactly as shown Parentheses must be placed exactly as shown Required keywords parameters and marks such as equal signs and parentheses are shown in bold type optional keywords parameters and marks are shown in regular type In the example KEYWORD req paraml req param2 OPT KEYWORD opt param KEYWORD must be entered exactly as shown req param1 must be replaced with a value variable or expression the equal sign must be entered req param2 must be replaced with a value variable or expression OPT KEYWORD can be omitted but must be entered exactly as shown if used opt param may be replaced with a value variable or expression but assumes a default value if not used The keyword type function program instruction and so on is shown at the top of the page An abbreviated syntax is shown for some keywords This is done when the abbreviated form is the most commonly used variation of the complete syntax Function This section gives a brief description of the keyword Usage Considerations This section lists any restrictions on the keyword s use If specific hardware or other options are required they will be listed here 1 In the SEE editor instructions can be abbreviated to a length that uniquely identifies the keyword The SEE editor will automatically expand the instruction to its full length 18 Adept Vision Reference G
54. in the previous example and find the edges using grayscale calculations with subpixel precision and put the signed gradient magnitudes in the array maggsl Draw the ruler so it can be seen dmode 1 along with the edges found Find only light to dark edges edge dir 1 VRULERI 2 1 2 xx mags 1 30 20 80 Related Keywords VDEEFAOI program instruction VFIND POINT program instruction VWINDOWI program instruction V BINARY system parameter V EDGE STRENGTH system parameter Adept Vision Reference Guide Rev A 219 VSELECT Program Instruction Syntax VSELECT mode ibr camera Function Select a virtual frame buffer for processing or display and optionally select a vir tual camera and its calibration to be associated with the frame buffer Parameters mode 0 select a virtual frame buffer 1 deselect a frame buffer see Details below ibr Real value variable or expression that specifies an image buffer region The last three digits of the image buffer region which specify a virtual frame buffer are used to select the virtual frame buffer The special values 1011 and 1012 select virtual frame buffers 11 and 12 which are full frame buffers In addition for convenience and backward compatibility the values 1 11 and 1001 select frame 11 and values 2 12 and 1002 select frame 12 camera Optional real value variable or expression that specifies a virtual camera number to be asso
55. initial search location In the examples shown below the guidelines are horizontal angle 0 so the search windows are rectangles and the width of the search window is the height of the rectangles The examples illustrate the three possible initial search locations The different possible positions for the initial search point allow control over the processing of multiple edges in the search window If the initial search point is in the center position edge points nearest the guideline are used when there are multiple edge points perpendicular to the guideline to choose from Otherwise the edge points nearest the side containing the initial search point are used See Figure 2 7 The VFIND LINE operator has a polarity that determines whether a light to dark or dark to light edge is being fit The vision system assumes that the dark side of the edge is on the top of the line being searched for and the light side is on the bottom assuming angle is 0 For visual reference one half of the search AdeptVision Reference Guide Rev A Program Instruction VFIND LINE window is drawn in dark blue and the other half is drawn in light blue the guideline is drawn in green and the initial search point is drawn in yellow the fit line is drawn in red and the point on the line closest to the initial search point is drawn in white See Figure 2 7 pos 0 pos 0 Edges Found pos 1 ang 180 pos 1 Figure 2 7 Line F
56. model The name is first checked against the lists of prototypes for backward compatibility Therefore you should not use the same name for a prototype model and an ObjectFinder model AdeptVision Reference Guide Rev A 243 VSTORE Monitor Command NOTE When no model names are specified VSTORE stores all the models in memory to the disk file There is no way to globally store only the ObjectFinder models If you want to do this you must store each model individually by specifying the model name It will assume all prototypes if no names are given This command saves the specified prototypes fonts or correlation templates in the given disk file Note however that prototypes fonts and templates may not be mixed in a file Thus all the model names must be prototypes or they must all be fonts or they must all be templates If no model names are specified all the prototypes currently in system memory are stored in the file As prototypes are stored their names are displayed in the Monitor display window along with their virtual camera associations When prototypes are stored any subprototype definitions are stored along with them As fonts or templates are stored the numbers of the fonts or templates are displayed in the Monitor window If the model name FONT_0 is specified all the defined fonts are stored in the file Similarly if the model name is TMPL_0 all the defined templates are stored As the file is crea
57. model up to 15 characters ibr Integer value specifying the image buffer region for training a model Image buffer regions specify both a size and a frame store see the description of VDEF AOI The image buffer region s Adept Vision Reference Guide Rev A 271 VTRAIN FINDER Program Instruction 272 AOI must specify a nonrotated rectangular shape For correlation templates the AOI is reduced if necessary to a multiple of four pixels value Floating point expressions whose meaning depends on the value of the mode parameter See below for details Description of the mode parameter This section describes the meanings of the different values used with the mode parameter and the meanings of other input parameters used with a particular mode If an input parameter is not shown it cannot be used for that mode Mode 1 Train first instance new model arg Hierarchical level 0 2 is the result of the subsampled image 0 Full resolution 1 One half resolution 2 One quarter resolution arg2 Verify percent arg3 Symmetry override Mode 2 Display features only arg Hierarchical level 0 2 is the result of the subsampled image 0 Full resolution 1 One half resolution 2 One quarter resolution arg2 Edges only 0 1 0 Features are fit to the edges and the features are displayed 1 Only the results of edge detection are displayed Mode 3 Set finder model parameters after last instance To train new instances during multi instanc
58. or 1 The 9 bits are combined into a single number ordered 987654321 1 is the least significant bit For instance consider a 3x3 binary neighborhood in the image as follows 1 1 1 1 1 0 1 0 0 When the above is combined into a single number the value in binary format would be 111110100 which is 500 decimal Now the VMORPH instruction operates as follows Let the 9 bit value for a neighborhood be called neigh If array neigh is not 0 the center pixel of the neighborhood is complemented Otherwise the center pixel is not changed Example The following sequence of program instructions defines a simple morphological operation to extract the boundary from a binary image The operation simply zeros binary pixels when the neighborhood is uniformly white all 1s and leaves all other pixels unchanged After executing this program the operation can be performed on an image by executing VMORPH 10 FOR i 0 TO 510 Zero almost the entire array morph i 0 END morph 511 1 This neighborhood causes a change VDEF MORPH 10 morph Define type 10 Related Keyword VMORPH program instruction 70 Adept Vision Reference Guide Rev A Monitor Command and Program Instruction VDEESUBPROTO Syntax VDEF SUBPROTO proto subname first_edge last_edge Function Define a subprototype Parameters proto proto is the name of the prototype of which the new subprototype subname is to b
59. or modify an Optical Character Recognition OCR font VDEFGRIP Define the shape and position of a robot gripper for clear grip tests VDEF LUT Define a user look up table for mapping graylevel values VDEF MORPH Define a binary morphological operation VDEF SUBPROTO Define a subprototype VDEF TRANS Define a transformation to apply to the location of all vision tools placed until the next VDEETRANS instruction VDELETE Delete a specified prototype subprototype Optical Character Recognition OCR font or correlation template in the vision system V DISJOINT Determine whether or not prototypes may be matched to multiple disjoint regions VDISPLAY Select the current vision display mode or the display mode to be used when the vision system performs its normal image processing functions V DRY RUN Enable simulation of various vision measurement and inspection operators VEDGE Compute edges in the grayscale image and threshold the edges replacing the binary image using either a cross gradient or Sobel algorithm VEDGE INFO Retrieve information about the edges and corners of a prototype or ofa region in the image V EDGE INFO Enable saving of information about edges in the image for recall via the VEDGE INFO instruction V EDGE STRENGTH Set the edge threshold for grayscale image processing and fine edge rulers V EDGE TYPE Determine the type of edge operator to use cross gradient or Sobel when a VPICTURE or VEDGE instruction is performed
60. process control information can be displayed continuously No display or no alter This mode is for high speed multiple camera applications A camera number should be specified along with this display mode Then when a VPICTURE is performed using that camera the current display mode is not changed For example if image from virtual camera 2 is being displayed in binary mode 2 the image from virtual camera 1 is being displayed in no alter mode 5 and a VPICTURE with virtual Adept Vision Reference Guide Rev A 81 VDISPLAY Monitor Command and Program Instruction camera 1 is issued the image from virtual camera 2 will continue to be displayed in binary mode If camera 0 all virtual cameras multiple VCOPY operations can be performed to the display frame store allowing you to create your own image high end systems only The overlay parameter to VDISPLAY may have the value 0 1 or 2 The overlay value 0 which is assumed when the system is initialized disables graphics overlay A value of 1 or 2 enables the overlay of the graphics memory over the image to be displayed as selected by the VDISPLAY mode 1 0 1 or 2 The source of the color overlay is the graphics frame store that is visible in VDISPLAY mode 3 Colors 1 through 16 are displayed as overlay colors An overlay value of 1 is like VDISPLAY mode 3 The vision display window is automatically erased when a VPICTURE operation is per
61. program instructions do not give any such notice When VPICTURE is executed as a monitor command several informational messages may be displayed These indicate which prototypes are known with their virtual camera associations and which require more training before they can be reliably recognized Execution of a VPICTURE instruction generally completes and thus the program continues after the frame grab is done but before all image processing is completed Consequently if the camera is mounted on a robot the controlling Vt program may have robot motion instructions following a VPICTURE instruction Then the vision system will process the acquired image while the robot is moving There are four exceptions to the statement that a VPICTURE instruction completes after the frame grab is done 1 If the V STROBE system switch is enabled the VPICTURE completes immediately after the strobe signal before the acquisition is done see the description of V STROBE for more details 2 If the VPICTURE mode is 2 quick frame grab and wait 0 VPICTURE completes after the acquisition has been started Not waiting until the acquisition is complete is appropriate only when the camera and scene are both still AdeptVision Reference Guide Rev A Monitor Command and Program Instruction VPICTURE 3 If the VPICTURE mode is 2 quick frame grab and wait 1 VPICTURE completes immediately before acquisition begins This mode is use
62. program instruction VGAPS program instruction V HOLES system switch VPICTURE monitor command and program instruction V RECOGNITION system switch VSUBPROTO program instruction VWINDOW program instruction 164 AdeptVision Reference Guide Rev A System Parameter V MAX AREA Syntax V MAX AREA camera Function Set the maximum area above which the vision system ignores regions Usage Considerations A change to this parameter takes effect when the next VPICTURE command or instruction or VWINDOW instruction is executed The effects of changes to this parameter are visible in the Vision display window only in a special display mode Regions larger than V MAX AREA will still appear in display modes 1 and 2 This is an array of parameters one for each virtual camera See the AdeptVision User s Guide for details on setting parameters Details Regions that contain more than V MAX AREA number of pixels are ignored by the vision system during object recognition clear grip testing and standard ruler measurements This parameter is less commonly useful than V MIN AREA A situation in which VMAX AREA would be useful however occurs when a fixture is in the scene that is larger than the object to be recognized or measured In that case time can be saved by setting V MAX AREA smaller than the area of the fixture region but larger than the object region because then the fixture boundary is not analyzed When a region
63. returned for each character in the expected text data i 5 1st character pick data i 6 Score of 1st character pick data i 7 2nd character pick data i 8 Score of 2nd character pick These four items repeat for each character region found and analyzed a total of data i 0 times This additional information may be used in the V program to permit certain alternates to the expected Operation 3 Recognition Recognition means there is no expected text Each character region in the image is compared to each character in the font For each character region found and analyzed the two top scoring characters in the font are returned along with their scores The data returned is almost identical to that described above data i 0 Number of character regions found and analyzed data i 1 Average score of all top picks data i 2 Minimum score of all top picks data i 3 0 data i 4 0 data i 5 1st character pick data i 6 Score of 1st character pick data i 7 2nd character pick data i 8 Score of 2nd character pick The last four items repeat for each character region found and analyzed a total of data i 0 times Regardless of the operation performed position information is returned when the array locs is specified in the VOCR instruction Returning this information takes additional time so if it is not needed the array 1ocs should not be specified When the array
64. roundness 4 PI area perim perim TYPE Region roundness roundness VLOCATE nam Locate next region in the image END Related Keywords VFEATURE real valued function V 2ND MOMENTS system switch V CENTROID system switch V HOLES system switch V MIN MAX RADII system switch AdeptVision Reference Guide Rev A 199 VPICTURE Syntax VPICTURE Function Monitor Command and Program Instruction camera wait acq_ibr sel_ibr mode how_many Acquire an image into a frame store and or initiate processing Usage Considerations The VPICTURE program instruction does not suspend program execution while the image is being processed In other words program execution continues with the next instruction while the vision system processes the image However if the parameter V IO WAIT is set to 1 image processing does not begin until the fast digital input interrupt signal is received Thus in that case program execution can be suspended for an indefinite time while the system waits for the high speed digital input signal NOTE Some vision instructions cause an error if they are executed while vision processing is still active The VWAIT program instruction will suspend program execution until the vision processor is idle Parameters camera wait acq_ibr sel_ibr 194 Optional real valued expression that specifies the virtual camera number The value 1 is assumed if no camera is specified Optional real valued e
65. subprototype or hole s region The subprototype numbers are relative to the prototype not the subprototype otherwise item 32 would always be 1 in get subproto mode ITEMS 34 to 36 After VSHOW only During prototype training constraints can be defined that limit recognition to a specific area and rotation relative to the defined prototype These items return the x y and angular constraints defined during training ITEM 40 After VLOCATE only This is the total area of all holes in the region last located with a VLOCATE instruction Adept Vision Reference Guide Rev A 101 VFEATURE Real Valued Function ITEM 41 After VLOCATE only If the V PERIMETER system switch has been enabled this is the outer perimeter of the region in millimeters ITEMS 42 and 43 After VLOCATE only If the VCENTROID system switch has been enabled these are the X Y coordinates of the region centroid These items are undefined if V CENTROID has not been enabled The positions and areas of holes are not considered in the centroid computation The VSUBTRACT HOLE system switch does not affect this ITEMS 44 to 47 After VLOCATE only If the switches V MIN MAX RADI and V CENTROID have been enabled these VFEATURE items indicate the points on the boundary of the region that are closest to and farthest from the region centroid Items 44 and 45 indicate the direction of the closest and farthest points relative to the region centroid Items 46 and 47 a
66. switch does not affect this However if you want the centroids and areas of holes you should enable V HOLES disable V DISJOINT and do VLOCATEs in get hole mode If prototypes are being recognized in the image V CENTROID does not have to be enabled to get the locations of the recognized objects Region centroids are different from recognized object locations although both are available with the VFEATURE function If region centroids are not needed V CENTROID should be disabled for efficiency If the V2ND MOMENTS or V MIN MAX RADII system switch is enabled the V CENTROID switch must also be enabled in order for the desired computation to be performed Example The following program segment draws a dot at the centroid of each region in the image ENABLE V BOUNDARIES V CENTROID Enable necessary switches VDISPLAY 3 Special display VPICTURE 0 Take a picture with virtual Camera 1 and no recognition ATTACH vlun 4 GRAPHICS Attach to the vision window FOPEN vlun Vision MAXSIZE 640 480 and select graphics scaling AdeptVision Reference Guide Rev A 39 V CENTROID System Switch GTRANS vlun 1 in real world millimeters v cx 42 Indexes of VFEATURE function v cy 43 for centroid VWAIT Wait for image processing to complete for graphics instr VLOCATE nam Locate anything in the image WHILE VFEATURE 1 DO cx VFEATURE vf cx cy VFEATUR
67. system switch is enabled all the candidate edge points within the search window that is those that have the correct polarity are displayed in white Note that displaying all of the edge points is computationally expensive so V SHOW EDGES should be enabled only for investigative purposes AdeptVision Reference Guide Rev A Program Instruction VFIND POINT Like VFIND LINE the VFIND POINT operator always searches for edges along horizontal vertical or diagonal directions Consequently the search window is a parallelogram not necessarily a rectangle The sides of the window that contain the initial search point can have any orientation but the other two sides of the search window are always horizontal vertical or diagonal VFIND POINT returns a coordinate whose distance from the side of the search window containing the initial search point o is accurate Note that the distance but not necessarily the coordinate is accurate In particular the coordinate has only about one pixel of accuracy along the axis parallel to the o side of the window The accuracy of the coordinate along the axis that is perpendicular to the o side of the window is much better particularly for fine edge VFIND POINT operations similar to that of a fine edge ruler The effort level specifies the fineness of the comb used in the search for edge points VFIND POINT searches inside the window along search lines that run perpendicular t
68. that specifies the disk file into which the models are to be stored This consists of an optional physical device an optional disk unit an optional directory path a filename and an optional file extension Uppercase or lowercase letters can be used The current default device unit and directory path are considered as appropriate see the DEFAULT command in the V Operating System User s Guide If no filename extension is specified the extension VS is appended to the name given Optional string expression that specifies the name of a current prototype OCR font or correlation template to be saved in the given file Uppercase or lowercase letters can be used If no model names are listed all prototypes are stored In this case the equal sign can be omitted Font names have the form FONT_n where n is the number of the font in the range 1 to 99 The special font name FONT_0 refers to all the fonts Template names have the form TMPL_n where n is the number of the template in the range 1 to 99 The special template name TMPL_0 refers to all the templates AdeptVision Reference Guide Rev A 245 VSTORE Program Instruction Details The type of model being stored is determined by the name which has been trained as either a prototype or an ObjectFinder model The name is first checked against the lists of prototypes for backward compatibility Therefore you should not use the same na
69. transformations implicitly contain the x and y millimeter per pixel scaling contained in the array scaler In addition the transformations may contain terms that correct for perspective distortion If the two transformations are not specified with VPUTCAL correction for perspective distortion is automatically disabled See Appendix C for more information on perspective distortion Example Calibrate virtual camera 2 using the data in cal array VPUTCAL 2 cal array Related Keyword VGETCAL program instruction Adept Vision Reference Guide Rev A 205 VPUTPIC Program Instruction Syntax VPUTPIC camera type zoom pic r c x0 y0 Function Store into a frame store an image saved previously with VGETPIC Usage Considerations The VISION switch must be enabled the vision processor must be idle and vision model training must not be active for this instruction to be executed Parameter camera Optional real valued expression that specifies the virtual camera number If not specified the camera number in the picture header at pic r c is used This parameter is currently ignored type Optional real valued expression indicating the type of data to expect in the pic array If not specified the value in the picture header is used 1 Grayscale image and binary or edge image 2 Binary or edge image only zoom Optional real valued expression indicating the zoom factor to use when writi
70. which color black or white is to be considered the background V BINARY camera Enable or disable automatic edge image generation at VPICTURE time V BORDER DIST camera Define an image border reduction in pixels to mask out regions clipped by the image border V BOUNDARIES camera Enable or disable boundary analysis by the vision system V CENTROID camera Enable computation of the centroid of each region in the image VCONVOLVE cam src_ibr Perform an image convolution on a grayscale frame type dmode dest_ibr possibly storing the result in a different frame store VCOPY cam scale dmode lut dest_ibr src_ibr Copy the image from one image buffer region to another AdeptVision Reference Guide Rev A VCORRELATE cam mode dmode max_depth accept give_up data i act_depth tplnum ibr VCORRELATE cam mode dmode data i tplnum shape cx cy dx dy ang Perform a normalized grayscale or binary correlation comparing a predefined template with a rectangular image window or searching for a closest match within the window to a predefined template VDEF AOI aoi shape diml dim2 dim3 dim4 angl ang2 Define an area of interest AOI Areas of interest are used by most vision tools to specify the tool placement within an image VDEF CONVOLVE type array i j Define an image convolution VDEF FONT op black_chars Define replace or modify an Optical Character Recognition
71. 1 and 0 typically take 60 80 milliseconds to complete depending on the complexity of the binary image assuming that the entire image is being processed as defined by the image buffer region Thus VPICTURE operations in mode 2 provide time savings for vision applications that use windows or inspection operators but no object recognition or region analysis for example centroids outside of windows For maximum speed of a VPICTURE in mode 2 enable the V BINARY system switch There are two grayscale physical frame stores that can acquire images While one frame store is acquiring a quick frame grab the other can be analyzed via inspection operators in a V program This means that images can be continuously acquired 30 frames per second and analyzed in a ping pong manner The acg_ibr parameter to VPICTURE indicates which frame store is to receive the new image The se1_ibr parameter indicates which frame store to select for subsequent instructions such as VWINDOWI VRULER VFIND LINE etc The VSELECT instruction performs the same function as the sel_ibr parameter to VPICTURE Since VSELECT is a separate instruction however a program can acquire two camera images into the two frame stores and then use VSELECT to quickly switch back and forth between them while performing analyses The values of the parameters V FIRST COL V FIRST LINE V LAST COL and V LAST LINE are used to clip the window to the image That is
72. 16 y scale Vertical scale factor mm pixel 17 cam model 0 Normal RS 170 camera Panasonic GP MF702 Sony XC 77 etc 1 Panasonic GP CD40 2 Reserved 3 MF 702 asynch reset mode noninterlaced full frame 30Hz 4 MF 702 asynch reset mode interlaced full frame 5 MF 552 asynch reset mode interlaced full frame 6 7 Reserved 8 15 Initially same as 0 204 AdeptVision Reference Guide Rev A Program Instruction VPUTCAL Table 2 5 Elements of VGETCAL VPUTCAL Scaler Calibration Array Continued Index Name Value or Meaning 18 pixel Index 1 8 of pixel compensation table used for filling in missing compensation pixels table 0 no pixel compensation 19 20 not used Must be 0 Several of the entries in the scaler array documented here are not actually used by the vision system but are generated by the Adept calibration utility programs and are used by other Adept software Those entries identified here only for completeness are Element 2 Calibration method Element 3 Link number Element 13 bit 1 HPS used flag The value for xy ratio must be in the range 0 5 to 1 5 If the x scale y scale supplied to VPUTCAL is not in the acceptable range the error Invalid vision X Y ratio is reported when the next VPICTURE operation is processed The arrays pmm to pix and pix to pmm contain two dimension homogeneous transformations These
73. 2 1 for each item in Table 2 2 Exceptions are shown below ITEMS 11 After VLOCATE only The model number is the number of the ObjectFinder model used in the planning list 1 10 ITEMS 12 After VLOCATE only Instance ID is a count of the number of objects recognized and regions not recognized These ID numbers start at 1 and count up ITEM 27 The number of features in the model VLOCATEd All of the line segments and circular arcs in the model are counted even if the weight after multiple instance training is zero Table 2 3 VFEATURE Function Data for ObjectFinder following VSHOW Index Value Units comment 1 Valid TRUE FALSE 2 X millimeters 3 Y millimeters 4 Z millimeters 5 RX degrees 6 RY degrees 7 RZ degrees 9 Verify percentage percentage 12 Min verify percent percentage 13 Min_X millimeters 14 Max_X millimeters 15 Min_Y millimeters 16 Max_Y millimeters 17 of pairs integer 19 Avg verify percent percent Indexes 12 19 and 20 are a set that is computed following multi instance training Adept Vision Reference Guide Rev A 103 VFEATURE Real Valued Function 104 Table 2 3 VFEATURE Function Data for ObjectFinder following VSHOW Index Value Units comment 20 Max verify percent percent 21 Hierarchical level 22 Parent number always 1 23 Virtual camera number 1 32 24 Effort le
74. 273 VTRAIN FINDER Program Instruction The following settings are suggested ENABLE V FIT ARCS vc May be disabled if the image consists of mostly lines See the description above for details V MIN LEN vc 10 See the description above for details V MAX PIXEL VAR vc 2 5 See the description above for details V EDGE STRENGTHI vc 9 See the description above for details 274 AdeptVision Reference Guide Rev A Program Instruction VTRAIN MODEL Syntax VTRAIN MODEL cam plan model_name Stext ibr VTRAIN MODEL cam plan model_name Stext shape cx cy dx dy ang Function Train on a vision model a correlation template or an Optical Character Recog nition OCR font For correlation this instruction defines the template For OCR this instruction trains the vision system to recognize characters in a font or causes the vision system to plan the recognition strategy for a fully trained font Usage Considerations The frame store currently selected or the specified image buffer region must contain a valid picture Otherwise an error results The VISION switch must be enabled the vision processor must be idle and vision model training must not be active for this instruction to be executed Font training is sensitive to camera calibration Before training fonts mount focus and securely fix your cameras and lenses Then calibrate the camera s that will be used to train and perform OCR The s
75. 3x3 diagonal edge 0 0 0 0 2 7 13 4 11 Adept Vision Reference Guide Rev A 0 0 48 0 3 13 20 13 3 0 5 14 0 0 30 0 2 7 13 10 7x7 horizontal edge 4 11 0 0 0 0 0 0 0 8 5x5 gaussian average O ON WON GO 0O O O O oo O ao OSOS O OG O a VD Ww N O 11 10 Program Instruction 11 7x7 diagonal edge 3 11 17 11 0 3 1 13 7x7 gaussian average U A Q 1 VD 0 O OA Q 1 4 9 11 D O O O 0O 1 U 0 O OA Q 0 O A Q 1 3 3 2 1 3 3 2 VD OA N OA Q 12 7x7 diagonal edge 1 15 16 Unity reserved for future ou u Oo O O 0 G G oo eo o 0 0 0 0 0 0 0 0 ooo O G G ou nde O O O O O AdeptVision Reference Guide Rev A 3 10 11 3 VDEFCONVOLVE 3 11 17 11 0 3 1 N 11 10 VD Dn N DB Q FR ray VD Ww N O 59 VDEFCONVOLVE Program Instruction 60 Convolutions numbered 17 to 32 can be defined by the user When you define a convolution the instruction parameter type must be in the range 17 to 32 and the parameter array must be defined The arrangement of the array elements can be visualized as shown in Figure 2 4 the figure assumes the starting array indices are array 1 1 When you are defining a convolution the terms should add up to about 128 If both positive and negative terms are used the sum of the positive ones an
76. 4 62 75 182 234 243 245 275 Overview of perspective distortion 308 P Perspective 307 308 310 calibration 139 308 distortion 116 124 205 213 218 307 308 309 310 311 312 comments on 218 correction for 116 124 205 213 218 308 309 312 distortion overview of 308 transformation 308 309 Perspectively corrected AdeptVision Reference Guide Rev A 319 Index plane 218 distorted arc 218 circle 218 Program instructions V EDGE INFO 90 VABORT 24 VADD 25 VAUTOTHR 28 VCONVOLVE 41 VCOPY 43 VCORRELATE 45 VDEFAOI 50 VDEFCONVOLVE 56 VDEFFONT 62 VDEELUT 67 VDEF MORPH 69 VDEFSUBPROTO 71 VDEFTRANS 73 VDEFGRIP 64 VDELETE 75 VDISPLAY 79 VEDGE 85 VEDGE INFO 87 VFIND ARC 106 VFIND LINE 116 VFIND POINT 124 VFINDER 114 VGAPS 134 VGET AOI 138 VGET TRANS 145 VGETCAL 139 VGETPIC 141 VHISTOGRAM 146 VLOAD 158 VLOCATE 160 VMORPH 180 VOCR 182 VPLAN FINDER 201 VPUTCAL 203 VPUTPIC 206 VRULERI 213 VSELECT 220 VSHOW 225 VSHOW MODEL 234 VSTATUS 240 VSTORE 245 VSUBPROTO 250 VSUBTRACT 253 VTHRESHOLD 260 VTRAIN 267 VTRAIN FINDER 271 VTRAIN MODEL 275 VWAIT 282 VWINDOW 284 VWINDOWB 287 VWINDOWI 290 Proposing prototype to image matches 304 Prototype recognition 301 algorithms 301 analyze a region during 12 disable 12 211 enable or disable 12 211 normal prototype 186 planning for 197 switches used only with 263 Publications related 8 Q Questions application 15 R Real valued function
77. 6 VQUEUE monitor command 208 real valued function 210 VRENAME monitor command 212 VRULERI program instruction 213 VSELECT program instruction 220 VSHOW monitor command 222 rogram instruction 225 VSHOW MODEL program instruction 234 VSTATUS monitor command 239 VSTATUS program instruction 239 VSTORE monitor command 243 program instruction 245 VSUBPROTO program instruction 250 VSUBTRACT program instruction 253 VTHRESHOLD program instruction 260 VTRAIN monitor command 265 program instruction 267 VTRAIN FINDER program instructions 271 VTRAIN MODEL program instruction 275 VWAIT program instruction 282 VWINDOW program instruction 284 VWINDOWB program instruction 287 VWINDOWI program instruction 290 WwW Waiting for high speed trigger 149 322 Adept Vision Reference Guide Rev A Adept User s Manual Comment Form We have provided this form to allow you to make comments about this manual to point out any mistakes you may find or to offer suggestions about information you want to see added to the manual We review and revise user s manuals on a regular basis and any comments or feedback you send us will be given serious consideration Thank you for your input NAME DATE COMPANY ADDRESS PHONE MANUAL TITLE AdeptVision Reference Guide Rev A PART NUMBER 00962 01300 PUBLICATION DATE August 1997 COMMENTS MAIL TO Adept Technology Inc Technical Publications Dept 11133 Kenwood Rd Cincinnati OH 45242
78. 93 302 Binary processing enabling disabling 33 Binary threshold setting second threshold 22 Boundary analysis enabling disabling 38 Index C Calibration 204 205 309 and position for font training 275 for prototype training 265 267 for training 271 arrays 140 203 309 camera 42 44 218 265 267 271 275 308 data 218 font training sensitivity to 275 prototype training sensitivity to 265 267 training sensitivity to 265 267 271 275 correction for perspective distortion 116 124 213 218 11 12 139 140 203 218 221 308 pass 308 read 308 method 204 205 procedure 139 programs 139 203 308 utility 205 status 204 target 309 values array of 203 virtual camera 12 220 X Y ratio in the 293 Calls service 14 Centroid data enabling collection of 39 Chain encode perimeters 303 Classify features 304 Compatibility 8 summary 314 Connectivity analysis 302 Contents of VGETPIC VPUTPIC header string 143 Conventions 8 data AdeptVision Reference Guide Rev A 317 Index Convolution matrix 60 Correlation binary 45 47 48 221 matches 47 matching 47 normalized 10 47 49 operation 48 results of the 46 score 46 template 10 11 13 14 46 48 75 76 82 156 158 234 235 243 244 245 272 275 276 277 315 number 46 tool 47 Cross gradient edge operator 94 Customer service assistance phone numbers 14 D Debugging vision dry run 84 Description of planning types 201 Description of
79. ACT HOLE Syntax V SUBTRACT HOLE camera Function Determine whether or not hole areas are to be subtracted from region areas Usage Considerations A change to this switch takes effect when the next VPICTURE command or instruction or VWINDOW instruction is executed This is an array of switches one for each virtual camera See the AdeptVision User s Guide for details on setting switches Details If VSUBTRACT HOLE is enabled region areas do not include the areas of any holes in the regions Otherwise region areas do include the hole areas This difference changes the effects of the values of V MIN AREA V MAX AREA V MIN HOLE AREA and the minimum maximum areas associated with prototypes Furthermore the areas reported by VQUEUE and VFEATURE are affected by the setting of this switch When operating in grayscale nonbinary mode an object region in the image consists of edges so most of the object area is considered to be a hole Therefore if you need to monitor the total area of an object when operating in grayscale mode it is useful to disable VSUBTRACT HOLE item B With this switch disabled any break in the edge would cause the hole area to merge with the background and would be immediately discounted in the area calculation item C With V SUBTRACT HOLE enabled the hole area would always be subtracted from the total area item A A small break in the edge may not cause a significant enough difference in the total ar
80. AM Sproto name of the prototype of interest OCAL data edge_num num_edges ocorn OCAL type AdeptVision Reference Guide Rev A Program Instruction VEDGE INFO Display the prototype or subprototype and use VFEATURE to determine the number of edges it has VSHOW 2 proto p 1 IF NOT VFEATURE 1 THEN TYPE Sproto is not a known sub prototype GOTO 100 END num_edges VFEATURE vf num bounds ocorn 1 0 Need some initial values ocorn 2 0 For each edge of the prototype use VEDGE INFO to retrieve information about the edge FOR edge_num 1 TO num_edges VSHOW monitor command VSHOW program instruction V EDGE INFO system switch Adept Vision Reference Guide Rev A VEDGE INFO data proto edge_num IF ocorn 1 lt gt data 2 OR ocorn 2 lt gt data 3 THEN IF edge_num lt gt 1 THEN TYPE New region or edge gap END TYPE Corner data 2 10 data 3 END IF data 0 0 THEN Its a LINE TYPE edge_num Line Wt data 1 ELSE It s an ARC IF data 0 gt 0 THEN Stype Convex ELSE Stype Concave END TYPE edge_num type arc S TYPE Center data 6 10 data 7 S TYPE Radius data 8 S TYPE Wt data 1 END TYPE Corner data 4 10 data 5 ocorn 1 data 4
81. APS VSHOW and VSUBPROTO The maximum number of prototypes and subprototypes counted together that may be loaded in the system at one time is 25 Of course an unlimited number may be stored on disk Adept Vision Reference Guide Rev A 71 VDEESUBPROTO Monitor Command and Program Instruction Examples The following examples compare the VDEFSUBPROTO monitor command with the program instruction If VDEFSUBPROTO is performed as a monitor command proto subname must be a nonquoted string like the VSHOW parameter If VDERSUBPROTO is used as a program instruction the prototype must be specified with a string expression VDEF SUBPROTO CASTING TOP 3 9 Monitor command VDEF SUBPROTO CASTING TOP 3 9 Program instruction Related Keywords 72 V EDGE INFO program instruction VGAPS program instruction VSHOW monitor command VSHOW program instruction VSUBPROTO program instruction Adept Vision Reference Guide Rev A Program Instruction VDEETRANS Syntax VDEF TRANS mode dx dy angle scale Function Define a transformation to apply to the location of all vision tools placed until the next VDEF TRANS instruction Usage Considerations The transformation applies only to the task in which it is defined Thus each task that requires a transformation must execute a VDEF TRANS instruction for that task This transformation applies to all tools regardless of whether they use the area of interest definitio
82. AdepiVision Reference Guide Version 12 1 Part Number 00962 01300 Rev A August 1997 cert 150 Rose Orchard Way San Jose CA 95134 USA Phone 408 432 0888 Fax 408 432 8707 AY Y Otto Hahn Strasse 23 44227 Dortmund Germany Phone 49 231 75 89 40 Fax 49 231 75 89 450 adept 41 rue du Saule Trapu 91300 Massy France Phone 33 1 69 19 16 16 Fax 33 1 69 32 04 62 technology inc 1 2 Aza Nakahara Mitsuya Cho Toyohashi Aichi Ken 441 31 Japan 81 532 65 2391 Fax 81 532 65 2390 The information contained herein is the property of Adept Technology Inc and shall not be re preci in whole or in part without prior written approval of Adept Technology Inc The in ormation herein is subject to change without notice and should not be construed as a commitment by Adept Technology Inc This manual is periodically reviewed and revised Adept Technology Inc assumes no responsibility for any errors or omissions in this document Critical evaluation of this manual by the user is welcomed Your comments assist us in prepara tion of future documentation A form is provided at the back of the book for submitting your comments Copyright 1992 1997 by Adept Technology Inc All rights reserved The Adept logo is a registered trademark of Adept Technology Inc Adept AdeptOne AdeptOne MV AdeptThree AdeptThree XL AdeptThree MV PackOne PackOne MV HyperDrive Adept 550 Adept 550 CleanRoom
83. E vf cy Draw a dot on the region centroid GARC vlun cx cy 1 VLOCATE nam END Related Keywords 40 V BOUNDARIES system switch VFEATURE real valued function V 2ND MOMENTS system switch V HOLES system switch VLOCATE program instruction V MIN MAX RADII system switch V PERIMETER system switch V SUBTRACT HOLE system switch If a region was found Get centroid Cx Cy Locate next region in image AdeptVision Reference Guide Rev A Syntax Program Instruction VCONVOLVE VCONVOLVE cam type dmode dest_ibr src_ibr Function Perform an image convolution on a grayscale frame possibly storing the result in a different frame store Usage Considerations The AdeptVision Enhanced VXL Interface option is required for convolutions larger than 3x3 On standard vision systems only the center 3x3 section of larger convolutions is used for both user defined and predefined convolutions Thus standard systems may use larger convolutions but the results may not be as expected see the description of VDEF CONVOLVE for a list of predefined convolutions With the AdeptVision Enhanced VXL Interface option VCONVOLVE will execute faster Parameters cam type dmode dest_ibr src_ibr Selects the threshold parameters to apply to the convolved result Integer value in the range 1 to 32 inclusive A value in the range 1 to 16 performs a prede
84. ETRANS program instruction Adept Vision Reference Guide Rev A 145 VHISTOGRAM Monitor Command and Program Instruction Syntax VHISTOGRAM dmode array index ibr Function Compute the histogram for a grayscale frame store Usage Considerations The VISION switch must be enabled the vision processor must be idle and vision model training must not be active for this instruction to be executed Parameters dmode Optional real valued expression specifying the display mode 1 to display the histogram or 0 not to display it This parameter defaults to 1 do display if no value is specified In that case the parentheses can be omitted array Optional array into which the values are placed Elements index to index 127 are filled with the pixel counts for each of the possible intensity values 0 to 127 index Optional integer value that identifies the first array element to be defined in array Zero is assumed if the index is omitted If a multiple dimension array is specified only the right most index is incremented as the values are assigned ibr Optional integer value that specifies the image buffer region area of interest and frame store to use The area of interest must have been defined with a VDEF AOI instruction The default frame is the current frame the one selected most recently via VPICTURE VSELECT or by a menu pick using the mouse Details VHISTOGRAM computes a histogram for an image currently i
85. F Name of the prototype to be trained For the monitor command the name of the prototype must be a string constant not surrounded by quotes For the program instruction however the prototype name may be a string variable constant including quotes or a string expression Integer value specifying the image buffer region for training a prototype Image buffer regions specify both a size and a frame store see the description of VDEF AOI The image buffer region s AOI must specify a rectangular shape Optional real valued expression indicating the shape of the window Currently the only choice is 1 for a rectangular window NOTE If any of the following four parameters cx cy dx and dy are specified all four parameters must be specified cx cy dy ang Real valued expressions specifying the center coordinate of the window in millimeters Real valued expression specifying the width of the window in millimeters Real valued expression specifying the height of the window in millimeters Optional real valued expression specifying the orientation of the window in degrees The default is 0 degrees AdeptVision Reference Guide Rev A Program Instruction VTRAIN value If mode 3 the instruction is changing a prototype parameter and this parameter supplies the new value If mode 2 the instruction is altering the training pull down menus and this parameter is a bit mask indicating whic
86. FEATS Enable the special display of features used for ObjectFinder recognition V SHOW GRIP Enable the special display of clear grip tests VSHOW MODEL Display a model either a correlation template or an Optical Character Recognition OCR font and return information about it or return information about all the defined templates or OCR fonts V SHOW RECOG Enable the special display of the objects recognized V SHOW VERIFY Enable the special display of the verification step in the recognition process VSTATUS Display vision system status information in the Monitor display window VSTATUS Return vision system status information in a real array VSTORE Store in a disk file selected or all vision prototypes and their subprototypes Optical Character Recognition OCR fonts or correlation templates V STROBE Enable the firing of a strobe light in synchronization with taking pictures VPICTUREs VSUBPROTO Determine the percentage of an edge or subprototype that was verified during recognition Also this instruction can have the prototype position refined based on only a subprototype or a single edge producing an adjusted location for the prototype VSUBTRACT Subtract two binary or grayscale images V SUBTRACT HOLE Determine whether or not hole areas are to be subtracted from region areas V SYNC STROBE Select synchronous or asynchronous firing of a strobe light when a picture is taken that is when a VPICTURE is executed VTHRESHO
87. GAIN to reduce the intensity range Similarly you should increase V OFFSET to shift the video data toward the right and decrease V OFFSET to shift it toward the left For good results you may have to repeat the procedure of taking a picture computing the new histogram and adjusting V GAIN and V OFFSET a few times V OFFSET must be assigned an integer value in the range 0 to 255 inclusive The parameter is set to 255 when the Vt and AdeptVision systems are loaded into memory from disk Adept Vision Reference Guide Rev A 189 V OFFSET System Parameter Example Shift the incoming video by 96 PARAMETER V OFFSET 96 Related Keywords VHISTOGRAM monitor command and program instruction V GAIN system parameter 190 AdeptVision Reference Guide Rev A System Switch V OVERLAPPING Syntax V OVERLAPPING camera Function Determine whether or not objects may be overlapping in the image Usage Considerations A change to this switch takes effect when the next VPICTURE command or instruction or VWINDOW instruction is executed If this switch is enabled the V TOUCHING system switch is automatically assumed to be enabled This is an array of switches one for each virtual camera See the AdeptVision User s Guide for details on setting switches Details Normally objects to be recognized in the scene are not overlapping so V OVERLAPPING should be disabled Object recognition is more efficient in this case The set
88. HISTOGRAM Compute the histogram for a grayscale frame store V HOLES Enable or disable the accounting of interior features in all objects V IO WAIT Enable the synchronization of taking pictures VPICTUREs with an external event that triggers the fast digital input interrupt line VISION Enable the entire vision system V LAST COL Set the number of the last column of frame store data to be processed V LAST LINE Set the number of the last line of frame store data to be processed V LAST VER DIST Enable an extra verification of prototype to image matches and specify the pixel tolerance to use when determining boundary coincidence VLOAD Load vision prototypes Optical Character Recognition OCR fonts or correlation templates from a disk file VLOCATE Identify and locate an object in the scene V MAX AREA Set the maximum area above which the vision system ignores regions V MAX PIXEL VAR During a VFIND LINE or VFIND ARC operation this parameter specifies the maximum pixel distance from the fit edge beyond which edge points may be filtered out During boundary analysis this parameter sets the maximum pixel deviation allowed when fitting lines and arcs to region edges V MAX SD Set the distance in units of standard deviation from the fit line or arc beyond which edge points should be filtered out AdeptVision Reference Guide Rev A 11 Chapter 1 AdeptVision Keyword Summary Table 1 2 AdeptVision Keyword Summary Continue
89. I 14 4 1 xx 1 230 340 20 10 15 Use a type 5 ring segment shape 2 window with the system parameters for virtual camera 1 to perform an inspection of a half eaten doughnut looking for debris edge points The center of the doughnut s circle is located at the point 25 2 37 The outer and inner radii are 50 millimeters and 18 millimeters respectively The doughnut is oriented so that its angular range is 0 to 180 degrees VWINDOWI 5 1 xx 2 25 2 37 50 18 0 180 Related Keywords 294 VDEF AOI program instruction VHISTOGRAM monitor command and program instruction VRULERI program instruction VWINDOW program instruction V 2ND THRESH system parameter V BACKLIGHT system switch V BINARY system parameter V THRESHOLD system parameter Adept Vision Reference Guide Rev A AdepWision Quick Reference V 2ND MOMENTS camera Enable computation of the best fit ellipse for each region in the image V 2ND THRESH camera Set a second threshold for use during binary image processing VABORT task_id Abort any active vision processing and all pending vision operations associated with the given task number VADD cam type dmode dest_ibr srcl_ibr src2_ibr Add two binary or grayscale images VAUTOTHR dmode start end array index ibr Determine good thresholds for binary image processing based on the gradients in a grayscale frame store V BACKLIGHT camera Define
90. If a template is being deleted the model name has the form TMPL_n where n is the number of the template in the range 1 to 99 For the monitor command this parameter must be a string constant not surrounded by quotes For the program instruction however the name can be specified with a string constant including quotes or an expression The type of model being supported is determined by the name which has been trained as either a prototype or an ObjectFinder model The name first is checked against the lists of prototypes for backwards compatibility Therefore you should not use the same name for a prototype model and an ObjectFinder model VDELETE deletes a prototype that has been trained or loaded into the vision system It does not affect a prototype that has been stored in a disk file For the VDELETE monitor command the prototype or subprototype is displayed in the Vision display window and the operator is asked for confirmation before it is deleted AdeptVision Reference Guide Rev A 75 VDELETE Monitor Command and Program Instruction VDELETE deletes OCR fonts and correlation templates in a similar manner Examples Note that when VDELETE is used as a monitor command as in the first example below the model name is not enclosed in quotes When VDELETE is used as a program instruction however quotes must be included if a string value is specified Monitor command to delete the subprototype fillet
91. LD Threshold a grayscale image producing a binary image V THRESHOLD Set the camera grayscale value that separates black pixels from white pixels V TOUCHING Determine whether or not objects may be touching in the image and still be recognized VTRAIN Initiate training of the object whose name is specified VTRAIN FINDER Initiate training of the finder model whose name is specified Adept Vision Reference Guide Rev A 13 Chapter 1 How Can I Get Help Table 1 2 AdeptVision Keyword Summary Continued Keyword Description VTRAIN MODEL Train on a vision model a correlation template or an Optical Character Recognition OCR font For correlation this instruction defines the template For OCR this instruction trains the vision system to recognize characters ina font or causes the vision system to plan the recognition strategy for a fully trained font VWAIT Delay program execution until processing of a VPICTURE or VWWINDOW operation is complete VWINDOW Perform processing blob finder prototype recognition or ObjectFinder within a rectangular window in the image VWINDOWB Extract basic image information from within a rectangular window VWINDOWI Extract image information from within a window with any of the following shapes rectangle circle pie cut ring or ring segment How Can I Get Help The following section tells you who to call if you need help W thin the Continental United States
92. LVE program instruction 41 VCOPY program instruction 43 VCORRELATE program instruction 45 VDEF AOI program instruction 50 VDEF CONVOLVE program instruction 56 VDEF FONT program instruction 62 VDEF LUT program instruction 67 VDEF MORPH monitor command and rogram instruction 69 VDEF SUBPROTO monitor command and program instruction 71 VDEF TRANS program instruction 73 VDEFGRIP program instruction 64 VDELETE monitor command and program instruction 75 VDISPLAY monitor command and program instruction 79 VEDGE program instruction 85 VEDGE INFO program instruction 87 Verify match 305 VFEATURE function data 96 VFEATURE real valued function 96 VFIND ARC program instruction 106 VFIND LINE program instruction 116 VFIND POINT program instruction 124 VFINDER program instruction 114 VGAPS program instruction 134 VGET AOI program instruction 138 VGET TRANS program instruction 145 VGETCAL program instruction 139 VGETPIC program instruction 141 VHISTOGRAM program instruction 146 Virtual frame store 54 Vision coordinate systems 310 VISION system switch 151 Vision system switch 151 VLOAD monitor commmand 156 program instruction 158 VLOCATE program instruction 160 VMORPH program instruction 180 VOCR program instruction 182 VPICTURE monitor command 194 program instruction 194 VPICTURE monitor command and rogram instruction 194 VPLAN FINDER monitor command and program instruction 201 VPUTCAL program instruction 203 VPUTPIC program instruction 20
93. NT and do VLOCATEs in get hole mode The best fit ellipse is computed as follows The ellipse is centered on the region centroid The major axis of the ellipse is aligned with the region axis of least inertia The minor axis of the ellipse is perpendicular to the major axis The radii of the ellipse axes are chosen so that the ellipse area and region area are equal and their ratios of major minor axes of inertia are equal Example The following program segment draws a cross on each region in the image showing the axes of the best fit ellipses 20 Adept Vision Reference Guide Rev A System Switch ENABLE ATTACH FOPEN GTRANS v V BOUNDARIES V CENTROID 4 GRAPHICS Vision MAXSIZE 1 vlun vlun vlun CX 42 Ey 43 major T 640 480 axis 48 49 50 major radius h FH FH Ph Fh lt lt lt lt minor radius VDISPLAY 3 VPICTURE 0 VWAIT VLOCATE 9 Snam WHILE EATURE 1 DO VFEATURE vf cx VFEATURE vf cy VFEATURE vf major radius VFEATURE vf minor radius R 4 VFEAT E vf major axis SIN an LINE Y vlun cx dx cy dy cx dx ang 90 COS ang min SIN ang min LINE cx dx vlun cytdy cx dx OCATE O Snam END Related Keywords VFEATURE real valued function V BOUNDARIES syste
94. PICTURE cam Take a picture using virtual camera cam VWAIT Wait until vision processor is idle Related Keywords VDEEAOI program instruction VSTATUS monitor command VSTATUS program instruction AdeptVision Reference Guide Rev A 283 VWINDOW Program Instruction Syntax VWINDOW cam type dmode how_many ibr VWINDOW cam type dmode how_many shape cx Cy width height ang Function Perform processing blob finder prototype recognition or ObjectFinder within a rectangular window in the image Usage Considerations The VWINDOW program instruction does not suspend program execution while the image is being processed In other words program execution continues with the next instruction while the vision system processes the window image NOTE Some vision instructions cause an error if they are executed while window processing is still active You can use the VWAIT program instruction to suspend program execution until the vision processor is idle Adept recommends the first syntax Parameters cam Optional real valued expression that specifies a virtual camera number see below The default is 1 type Optional real valued expression indicating the type of the window Currently the only choice is 1 dmode Optional real valued expression specifying the display mode to use when displaying the border of the window 1 no draw 0 erase 1 draw solid 2 complement 3 draw dashed The default is 1 draw s
95. PIVIEW lt gt cerda RA 302 Connectivity Analysis 302 Chain Encode PerimeteIS 303 Ft Primitive EdgestoChains 2 ee ees 303 Ft Lines and ArcstoEdges 40 4 303 Classify Features a 304 Propose Prototype to image Matches 304 Vey Match l a aaa 0 305 Perspective Distor on 0 307 Overview of Perspective Distortion 308 AdeptVision Reference Guide Rev A 3 Table of Contents Calibration AMayS o 309 How to Use the Arays pmm to mmL and mm to pmm 311 D Upgrading ProgramCode 313 WHVOCUCUON 5 5 coe e arda AAA 314 Compatibility Summary 2 2 ee 314 DAEK o oe Se he ae ee HH AE RRR oS ok ae eg SS a aT 317 4 AdeptVision Reference Guide Rev A Figure 2 1 Figure 2 2 Figure 2 3 Figure 2 4 Figure 2 5 Figure 2 6 Figure 2 7 Figure 2 8 Figure 2 9 Figure C 1 List of gures Example of V BORDERDIST 2 37 Shape Parameters for Rectangular Tools 51 Shape Parameters forArc Shaped Tools 53 Arrangement of Elements of Convolution Matix 60 EffectofV DISOINTSwitth 78 Arc FinderShape 048 111 Line Finder Tool Start Position and Polarity 119 Sample Line FinderTool 120 Effects of V MAX PIXELVARParameter
96. PIXEL VAR which is the pixel variance allowed when fitting the edges lines and arcs in the image With the VSUBPROTO instruction the verified percentages of edges of subprototypes or an individual prototype may be determined VSUBPROTO may also be used to refine the position of the recognized prototype based on a subprototype or a single prototype edge In this latter case the V ILAST VER DIST parameter may be assigned a value larger than VMAX VER DIST in order to reach out to edges that are suspected of deviating in position Like VMAX VER DIST V LAST VER DIST is in pixel units V LAST VER DIST must be assigned a real value in the range 0 to 16 inclusive The parameter is set to 0 when the V and AdeptVision systems are loaded into memory from disk That is the extra verification is disabled 154 Adept Vision Reference Guide Rev A System Parameter V LAST VER DIST Examples PARAMETER V LAST VER DIST vcam 0 PARAMETER V LAST VER DIST 2 PARAMETER V MAX VER Related Keywords VGAPS program instruction VSUBPROTO program instruction V MAX VER DIST system parameter AdeptVision Reference Guide Rev A DI sT 1 155 VLOAD Monitor Command Syntax VLOAD file spec Function Load vision models ObjectFinder models prototypes Optical Character Recog nition fonts or correlation templates from a disk file Usage Considerations VLOAD will not load files created prior to version 11 0 Parame
97. RECOG Related Keyword V RECOGNITION system switch AdeptVision Reference Guide Rev A 237 V SHOW VERIFY System Switch Syntax V SHOW VERIFY camera Function Enable the special display of the verification step in the recognition process Usage Considerations A change to this switch takes effect when the next region is displayed in the Vision display window with a graphical display mode in effect This is an array of switches one for each virtual camera See the AdeptVision User s Guide for details on setting switches Details This switch controls a diagnostic tool When the switch is enabled all attempts to verify proposed matches are drawn in the Vision display window overlaying the displayed regions That is as the lines and arcs of the prototype are compared to the edges in the image they are drawn in the Vision display window The verified parts are displayed in one color and the unverified parts in a different color Usually when a verification fails the proposed match fails not all of the prototype boundary is drawn This simply illustrates the early cutoff of the verification process That is the amount of the prototype boundary that did not verify exceeded the maximum allowed according to the prototype edge weights and verify percentage threshold NOTE This display option is often very time consuming The parameter V MAX TIME may have to be increased if the verification of all match proposals is to
98. Record endpoint of this edge ocorn 2 data 5 END 100 RETURN END Related Keywords 89 V EDGE INFO System Switch Syntax V EDGE INFO camera Function Enable saving of information about edges in the image for recall via the VEDGE INFO instruction Usage Considerations The VBOUNDARIES system switch must be enabled in order for edge information to be computed Also the V FIT ARCS switch must be enabled if circular arcs are to be fit in addition to lines A change to this switch takes effect when the next VPICTURE command or instruction or VWINDOW instruction is executed This is an array of switches one for each virtual camera See the general description of switches in the AdeptVision User s Guide for syntax details Details When regions are analyzed and arcs and lines are fit to the boundaries of the regions the information is stored away if V EDGE INFO is enabled The instruction VEDGE INFO may then be used to retrieve descriptions of the lines and arcs that were fit If the V FIT ARCS system switch is disabled no arcs will be fit Furthermore if the V MAX PIXEL VAR system parameter is 0 neither lines nor arcs will be fit to region boundaries In that case the VEDGE INFO instruction will return the primitive edges that bound the regions The time spent by the vision system fitting lines and arcs will be saved but usually many more primitive edges will then be present VEDGE INFO must then be executed more t
99. SHOW VERIFY system switch AdeptVision Reference Guide Rev A 83 V DRY RUN System Switch Syntax V DRY RUN Function Enable graphics only mode for various vision operators Usage Considerations A change to this switch takes effect when the next vision operator is executed Details This switch is for use by V programs that wish to display a vision operation such as VFIND LINE in the Vision display window and move it around without it actually fitting lines and showing results For example VisionWare uses this option extensively The setting of V DRY RUN affects all instructions that use an image buffer region When V DRY RUN is enabled tool graphics are displayed if the tool s display mode is set appropriately but the operation is not actually performed Example Enter graphics only mode ENABLE V DRY RUN 84 AdeptVision Reference Guide Rev A Program Instruction VEDGE Syntax VEDGE cam type dmode dest_ibr src_ibr Function Compute edges in the grayscale image and threshold the edges replacing the binary image using either a cross gradient or Sobel algorithm Parameters cam Optional integer value specifying a virtual camera The value of V EDGE STRENGTH associated with the virtual camera will be used to compute edges The default is 1 type Optional integer value indicating the type of edge algorithm to use 1 for cross gradient and 2 for Sobel The default is 1 cross gradient dmode Optio
100. STRACT Initiate font training on an input text string The font must have INPUT PARAM Sstr the text to train on 280 AUTO ang cam cx cy dx dy font vlun AUTO Sans Sfont LOCAL data Define these for your specific needs cam 1 Virtual camera number font 1 Font number 1 to 99 cx 256 Center of the window cy 242 dx 350 Width of the window dy 50 Height of the window ang 0 Orientation of the window model ibr 3001 Define aoi variable VDEF AOI model ibr 1 cx cy dx dy ang Sfont FONT SENCODE 10 font Compose name of the font VDISPLAY cam 2 1 Display binary image with overlay VPICTURE cam 2 Do quick frame grab ATTACH vlun 4 GRAPHICS Attach to the vision window FOPEN vlun Vision MAXSIZE 640 480 GTRANS vlun 1 Select scaling in real world mm VWINDOWI data cx cy dx dy ang Adept Vision Reference Guide Rev A Program Instruction VTRAIN MODEL Cl Place sample text in window Press Enter to train type S any character and Enter to halt training Cl TYP TYP WHILE TRUE DO VDISPLAY cam 0 1 GTYPE vlun 50 100 str PROMPT Press Enter to train on str ans IF Sans lt gt GOTO 10 VPICTURE cam 2 VTRAIN MODEL cam Sfont Sstr model ibr END 10 RETURN END Related Keywords VDEF AOI p
101. TE If only one image buffer region is specified both image buffer regions default to the specified image buffer region If neither image buffer region is specified both default to the currently selected full frame which is selected with a VSELECT or VPICTURE instruction for example VPICTURE 1 1011 Details VCOPY copies one image buffer region to another and optionally zooms the image Both the grayscale and binary images are copied If the scale parameter is 0 5 the image is averaged with a 2x2 convolution The result is then subsampled and thresholded using the threshold parameters for the virtual camera specified in cam If the scale parameter is 2 the source frame store must be the display frame store src_ibr must have the form nnn009 After a frame is copied the destination frame assumes some attributes of the source frame In particular the virtual camera that was associated with the source frame becomes associated with the destination frame This affects subsequent measurement or inspection operations on the frame such as VRULERI or VWINDOW because conversion of pixel units to millimeters is determined by the camera calibration NOTE When copying from one area to another within the same frame buffer do not copy areas of interest that would overlap areas that will be copied out of later in the operation Specifically you cannot VCOPY a 100x100 pixel region down 10 pixels By the time the tenth row is read out f
102. TOUCHING V DISJOINT per Region per Scene Off Off 1 No limit Off On 1 how_many On Off how_many No limit On On how_many how_many Example Inform the vision system that objects may touch in the image ENABLE V TOUCHING Related Keywords 264 V BOUNDARIES system switch V DISJOINT system switch V OVERLAPPING system switch V RECOGNITION system switch Adept Vision Reference Guide Rev A Syntax Monitor Command VTRAIN VTRAIN prototype shape cx cy width height ang VTRAIN prototype ibr Function Initiate training of the prototype whose name is specified Usage Considerations The command and the instruction both require that all virtual cameras be idle that is no VPICTURE or VWWINDOW operation can be executing Prototype training is sensitive to camera calibration Before training prototypes mount focus and securely fix your cameras and lenses Then calibrate the camera s that will be used to train and recognize the prototypes The same camera s with the same calibration and position used to train a prototype must be used to recognize the prototype Parameters prototype shape Name of the prototype to be trained For the monitor command the name of the prototype must be a string constant not surrounded by quotes For the program instruction however the prototype name may be a string variable constant including quotes or a string expression Optional real valued expression indica
103. a i 3 X Y coordinates of the arc center in millimeters data i 4 Radius of the arc in millimeters data i 5 Percentage of the estimated arc s extent for which edge points were found 0 0 to 100 0 data i 6 Maximum error distance from the fit arc to the most distant edge point found in pixels data i 7 Maximum error toward the inside of the circle or arc in pixels data i 8 Maximum error toward the outside of the circle or arc in pixels data i 9 Percent of edge points filtered out Optional array index that identifies the first element to be defined in data The default is 0 If a multiple dimension array is specified only the right most index is incremented as the values are assigned Integer value specifying the image buffer region within which to search for an arc Image buffer regions specify both a size and a frame store see the description of VDEF AOD Real valued expressions specifying the center coordinate of the estimated arc circle in millimeters This coordinate may be outside the image but it must be within a 1000 pixel radius of the image origin bottom left corner after being transformed from millimeters to image pixels Real valued expression specifying the radius of the estimated arc circle in millimeters The value for r must be in the range 1 to 512 image pixels after it is converted from millimeters Furthermore there is a minimum arc circumference see ang0 angn
104. able for analysis Any VPICTURE mode may be used except future frame grab mode 1 Quick frame grab mode 2 is the recommended mode The information that VWINDOWI returns in its array parameter is determined by the type parameter The table below lists the value each array element holds for each window type the actual index for a data item is index plus the item s relative index shown AdeptVision Reference Guide Rev A 291 VWINDOWI Program Instruction Relative Array Index Window Type 0 1 2 3 4 5 6 7 8 9 0 clipped width ht area count 0 0 0 0 0 1 clipped width ht area avg 0 0 0 0 0 2 clipped width ht area avg min max 0 0 0 3 clipped width ht area avg min max object bkgd 0 4 clipped width ht area avg min max object bkgd stdev 5 clipped width ht area ecount 0 0 0 0 0 6 128 histogram pixels counts see below where clipped TRUE if the window extended past the edge of the image FALSE otherwise width ht The actual pixel width and height of the window after millimeter to pixel conversion and compensation for any nonsquareness of pixels area The actual number of pixels considered in the window If sample is 2 this value is close to 1 4 the total number of pixels in the window count The number of nonzero pixels in the window of the binary frame store as seen in VDISPLAY 2 This is only for type 0 windo
105. ach grayscale frame Define AOIs insp_ibr2 VDEF AOI insp_ibrl 1 380 240 50 50 O VDEF AOI insp_ibr2 1 150 150 100 10 0 Prime the pump by grabbing into frame 2 VPICTURE cam2 0 insp_ibr2 2 WHILE TRUE DO Loop forever E Grab into 1 select 2 for a window inspection VPICTURE caml 0 insp_ibrl insp_ibr2 2 VWINDOWI cam2 5 1 data insp_ibr2 IF data 4 lt 250 OR data 4 gt 350 THEN Report deviation in edge counts here END Grab into 2 select 1 for a window inspection VPICTURE cam2 0 insp_ibr2 insp_ibrl 2 VWINDOWI caml 5 1 data insp_ibrl IF data 4 lt 100 OR data 4 gt 150 THEN Report deviation in edge counts here Related Keywords VDEF AOI program instruction VLOCATE program instruction VWINDOW program instruction VSELECT program instruction V BINARY system switch V DISJOINT system switch V OVERLAPPING system switch V TOUCHING system switch 200 AdeptVision Reference Guide Rev A insp_ibril 1011 saoi 1 virt frame buffer 11 2012 jaoi 2 virt frame buffer 12 Monitor Command and Program Instruction VPLAN FINDER Syntax VPLAN FINDER cam type dmode Sfmods bmods Function Set up the type of planning used by the Finder when locating models Parameters cam Real valued expression indicating the virtual camera number to use for associating planning with and t
106. ach virtual camera See the AdeptVision User s Guide for details on setting parameters Details Features that contain fewer than V MIN LEN number of pixels are ignored by the vision system during feature pairing operations This parameter should be set large enough so that the vision system does not waste time analyzing noise spots but small enough so that all features of a minimum desired length are included This parameter must be assigned an integer value there is no minimum or maximum value The default value is 40 pixels Exa mple Ignore features that are smaller than 25 pixels in length PARAMETER V MIN LEN 25 AdeptVision Reference Guide Rev A 177 V MIN MAX RADII System Switch Syntax V MIN MAX RADII camera Function Enable the feature that for each region in the image finds the two points on the perimeter that are closest to and farthest from the region centroid Usage Considerations The system switches VBOUNDARIES and V CENTROID must be enabled in order for VMIN MAX RADII to have its effect A change to this switch takes effect when the next VPICTURE command or instruction or VWINDOW instruction is executed This is an array of switches one for each virtual camera See the AdeptVision User s Guide for details on setting switches Details When the switches V MIN MAX RADII and V CENTROID are both enabled the perimeters of all regions are scanned to find the points closest to and farthes
107. adius of the image origin bottom left corner after being transformed from millimeters to image pixels Ruler type 1 returns the graylevel values at each pixel along the path of the ruler The first element in the array is the count of pixels along the ruler The starting color is always 0 Graylevel rulers can be used to determine the average graylevel along a line The maximum length of a type 1 ruler is 508 pixels If a linear ruler starts outside the frame store the off frame pixels are padded with zeros If the end of a linear ruler goes outside the frame store no padding is done In that case the array is just shorter If any part of a graylevel type 1 arc ruler falls outside the frame store no information is returned other than an indication that the ruler was clipped Edge rulers do not report edges that occur within one pixel two pixels for arc rulers of the edge of the image the start of the ruler or the end of the ruler This is because the information needed to compute the edge position is not complete in these areas The time required to process a ruler depends on the orientation and length of the ruler and the number of transitions encountered The parameter maxcnt may be used to limit the latter In addition the speed is affected if you are having the ruler displayed or are getting grayscale edge magnitudes returned All these things affect the speed of each ruler type differently 1 The VWINDOWT instruct
108. alers i pmm to pix j k pix to pmm 1 m to cam Load vision calibration data from 1 to 3 arrays VPUTPIC camera type zoom pic r c x0 y0 Store into a frame store an image saved previously with VGETPIC VQUEUE camera Display object information for any objects queued in the vision system awaiting retrieval by VLOCATE instructions VQUEUE camera name Return the number of objects in the vision system queue V RECOGNITION camera Enable or disable prototype recognition by the vision system VRENAME new_name old_name Rename a prototype or subprototype VRULERI cam type edge dir data i mags j ibr VRULERI cam type dmode maxcnt edge dir data i mags j 1 x0 y0 len dmode maxcnt ang VRULERI cam type dmode maxcnt edge dir data i mags j shape cx cy radius ang0 angn Obtain edge information or graylevels along a line or circular arc in the current image VSELECT mode ibr camera Select a virtual frame buffer for processing or display and optionally select a virtual camera and its calibration to be associated with the frame buffer VSHOW proto_name grip edge_num List the defined prototypes or display a vision prototype a subprototype or a specific prototype edge in the Vision display window Edge numbers are optionally shown VSHOW mode edge_num Display a vision model ObjectFinder prototype or subprototype and make information about it proto_name tr
109. ally In addition to external encoders for example those on conveyor belts motor encoders with the AdeptMotion VME option and the joint encoders on the Adept robot are latched when the strobe light fires The latched encoder positions can be accessed with the V DEVICE function For version 10 x and older If an object in the image has been VLOCATEd the position of external encoder 1 is also available as VFEATURE 8 The position of a belt encoder may be used with the V belt tracking features to locate and pick up objects from a moving belt The V LATCH and PLATCH functions return the latched robot position as a transformation and as a precision point respectively See the V Language Reference Guide for details on the DEVICE LATCH LATCHED and PLATCH functions Example Enable the strobe light for virtual camera 1 ENABLE V STROBE 1 Related Keywords VPICTURE monitor command and program instruction V IO WAIT system parameter V SYNC STROBE system parameter Adept Vision Reference Guide Rev A 249 VSUBPROTO Program Instruction Syntax VSUBPROTO ver_percent trans_var subproto_name edge_num Function Determine the percentage of an edge or subprototype that was verified during recognition Also this instruction can have the prototype position refined based on only a subprototype or a single edge producing an adjusted location for the prototype Usage Considerations The VISION switch must be enabled t
110. am instruction The perspective transformation is a two dimensional homogeneous transformation that contains the scaling described above plus correction for perspective distortion Furthermore the transformation may contain a small translation and rotation Note that the scale transformation is a simplification of the perspective transformation The Adept camera calibration programs including the built in VisionWare calibration perform calibrations that correct for perspective distortion After a calibration these programs tell the user how much perspective distortion is present in the image and the direction the image plane is tilted so that the user can adjust the camera orientation or the workspace to compensate for the distortion The user can elect to disable correction for perspective distortion in order to maximize run time speed If enabled perspective calibration is used to correct the results of linear and arc rulers and line point and arc finders All other vision tools return results based on the scale transformation The two V instructions that deal with vision calibration are VGETCAL and VPUTCAL Normally you will never need to use these instructions directly The calibration programs supplied by Adept pass calibration data to the vision system with the VPUTCAL instruction The load area subroutine may be included in your application program to 1 read calibration data previously saved in a disk file and 2 perfor
111. am instruction VABORT monitor command and program instruction VWAIT program instruction V STROBE system switch V SYNC STROBE system parameter 1542cm per 512 pixels 0 01 cm pixel 0 1 mm pixel 2 0 016 second 1000 mm second 0 1 mm pixel 160 pixels 150 AdeptVision Reference Guide Rev A System Switch VISION Syntax VISION Function Enable the entire vision system Usage Considerations This switch must be enabled before any vision commands or instructions may be executed Details The vision processor initializes itself when the VISION system switch is enabled The vision processor automatically enables VISION on power up If the switch is already enabled and a picture is being processed a VPICTURE or VWINDOW has been issued earlier typing ENABLE VISION makes the vision system stop processing images and reinitialize itself When VISION is enabled the vision system performs an equipment check If there is a failure the V system displays an error message in the Monitor display window such as Camera interface board absent When the VISION switch is disabled the vision system closes the vision display window removing the menu selections When enabled the vision system reopens the window and restores the menu selections assuming that a user task does not have the window opened for read write Example Restart the vision system ENABLE VISION AdeptVision Reference Gu
112. ame camera s with the same calibration and position used to train a font must be used to recognize characters in the font Correlation templates are independent of camera calibration but are sensitive to the image size which can change by moving the camera changing lenses or refocusing a lens Make sure the camera setup remains unchanged during template training and matching Adept recommends the first syntax Parameters cam Optional real valued expression indicating the virtual camera number to use for selecting various parameters for training an OCR font such as V THRESHOLD and V MIN AREA The default camera is 1 plan If smodel_name specifies an OCR font optional real valued expression that specifies the operation desired When planning is requested only the plan and model_name parameters are considered All the others can be omitted 0 Directs the vision system to perform OCR training Adept Vision Reference Guide Rev A 275 VTRAIN MODEL 276 model_name Stext ibr shape Program Instruction 1 Directs the system to plan recognition and to display associated graphics see below 1 Directs the system to plan recognition but without display of associated graphics see below If Smode1_name specifies a correlation template s mode of training as follows O Default creates a hierarchical grayscale template 1 No longer used 2 Creates a nonhierarchical grayscale template S
113. and and Program Instruction 196 The vision system queues one for each virtual camera have a specified amount of memory allocated for recognized and unrecognized regions This allocation can be set with the DEVICE program instruction Holes are included in the queues if the V HOLES system switch is enabled and the V DISJOINT system switch is disabled An image is processed when a VPICTURE operation is performed in mode 1 or 0 and all the object data for a particular virtual camera is retained in that virtual camera s queue until another VPICTURE or VWWINDOW operation for the same virtual camera is executed This means that the user may accumulate located object information gathered from several different virtual cameras before accessing it with VLOCATE instructions Keep in mind that a significant time delay perhaps as long as a minute or more can occur before image processing commences This delay occurs the first time a VPICTURE or VWINDOW operation is executed for a particular virtual camera after prototype training VTRAIN has been performed or certain recognition parameters and switches have been altered for that virtual camera During this interval the vision system analyzes the trained information to generate the data that it requires to perform the actual image processing If a VPICTURE monitor command causes this planning the message Planning please wait is displayed in the Monitor display window The VPICTURE and VWINDOW
114. and dim4 If both these dimensions are negative dim1 and dim2 specify the upper right corner rather than the lower left corner NOTE The area of interest widths are automatically reduced to a multiple of four pixels a dim3 gt Shape 1 y lt dim3 p ei EE lt angi dim1 dim2 pa dim1 dim2 lt dim3 lt angi dim1 dim2 lt dim3 lt angi O dim1 dim2 dim1 dim2 dim3 dim4 dim1 dim2 Shape 4 A dim3 dim4 dim1 dim2 Figure 2 2 Shape Parameters for Rectangular Tools Adept Vision Reference Guide Rev A 51 VDEFAOI Program Instruction 52 Figure 2 3 shows the shape parameters that define arc shaped tools As in rectangle shaped tools a fully defined area of interest can be used by simpler shaped tools and the appropriate dimensions are ignored In the illustration dim1 and dim2 always refer to the center of the arc circle For shape 7 a positive value for ang2 indicates a counterclockwise range from ang and a negative value indicates a clockwise range The sign of ang2 also determines whether ruler tools search for edges in a clockwise or counterclockwise direction Shape 3 is inappropriate but allowed for finder tools Shape 4 is not allowed for line or point finders Shapes 9 and 10 are inappropriate but allowed with the VFIND ARC instruction AdeptVision Reference Guide Rev A Program Instruction VDEEAOI Shape 5 Shape 6 Shape 7 Shap
115. ans_var grip 297 V SHOW BOUNDS Appendix A AdeptVision Quick Reference available through the VFEATURE real valued function V SHOW BOUNDS camera Enable the special display of the lines and arcs fit to the boundaries of regions V SHOW EDGES camera Enable the special display of edges both the primitive edges that are fit to the boundaries of regions and the edge points that are found by the finders VFIND LINE VFIND ARC and VFIND POINT V SHOW FEATS camera Enable the special display of features used for ObjectFinder recognition V SHOW GRIP camera Enable the special display of clear grip tests VSHOW MODEL mode Schars model_num Display a model either a correlation template or an Optical Character Recognition OCR font and return information about it or return information about all the defined templates or OCR fonts data i V SHOW RECOG camera Enable the special display of the objects recognized V SHOW VERIFY camera Enable the special display of the verification step in the recognition process VSTATUS Display vision system status information in the Monitor display window VSTATUS camera type array index Return vision system status information in a real array VSTORE file_spec model_name model_name Store in a disk file selected or all vision prototypes and their subprototypes Optical Character Recognition OCR fonts or correlation templates VSTORE lun
116. arc are displayed in white Edge points discarded during the filtering process are displayed in gray Note that displaying the edge points is computationally expensive so it should be enabled only for investigative purposes Items data i 7 data i 8 and data i 9 are returned only when edge filtering is enabled that is when V MAX SD is nonzero Item data i 9 is the percentage of edge points filtered out This percentage ranges from 0 up toward 100 It can never be 100 however because the filtering process always stops before filtering out too many points The execution time for VFIND ARC is proportional to the effort level multiplied by the circumference of the arc Also the execution time increases as the search range increases An approximate formula for computing the execution time in microseconds is shown below where C is the arc circumference in pixels E is the effort level and RR is the radius range in pixels This formula does not account for edge filtering which will add additional time that depends on the number of filtering iterations performed Time C E 11 RR 6 After an arc has been fit that is after the final fit if edge filtering is enabled the edge points used to fit the arc are compared with the fit arc and the distance from the fit arc to the most distant edge point is returned data i 6 This distance is in pixels If this value is not close to one the arc edge in the imag
117. associated with the prototype casting VDELETE casting fillet Program instruction to delete the model named by the string variable model VDELETE Smodel Program instruction to delete font 2 VDELETE FONT_2 Program instruction to delete template 1 VDELETE TMPL_1 Related Keywords VDEEFONT program instruction VDEESUBPROTO monitor command and program instruction VLOAD monitor command VLOAD program instruction VTRAIN monitor command and program instruction VTRAIN MODEL program instruction 76 AdeptVision Reference Guide Rev A System Switch V DISJOINT Syntax V DISJOINT camera Function Determine whether or not prototypes may be matched to multiple disjoint regions Usage Considerations A change to this switch takes effect when the next VPICTURE command or instruction or VWINDOW instruction is executed When this switch is enabled the V HOLES switch has its effect but not its setting disabled This is an array of switches one for each virtual camera See the general description of switches in the AdeptVision User s Guide for syntax details Details If V DISJOINT is enabled the system finishes processing the boundaries of all regions in the image before attempting object recognition Disjoint regions in the image may then contribute to the recognition of a single prototype When V DISJOINT is disabled each region in the image is processed to compl
118. at is verify that an instance of the prototype is visible in the image the vision system searches for image boundaries that are nearly coincident with the proposed positions of the prototype boundaries The system parameter V MAX VER DIST defines in terms of pixels the tolerance of the nearly coincident test The portions of the prototype boundaries that are located within V MAX VER DIST pixels of image boundaries are considered verified The parameter V LAST VER DIST is used in the same way as V MAX VER DIST but only during the final verification of each match Depending on the complexity of the prototype being recognized one or more verifications are performed per match proposal When a prototype to image match is verified all the edges of the prototype are compared to the nearby image edges that is edges within V MAX VER DIST pixels and statistics are gathered describing the deviations in position and rotation Following each verification the vision system computes a prototype to image transformation that minimizes the deviations in edge positions and rotations These steps of verification and repositioning may be performed a number of times in order to refine the fit of the prototype If the parameter V LAST VER DIST has a nonzero value an extra final verification is performed using V LAST VER DIST as the maximum verify distance V LAST VER DIST if smaller than V MAX VER DIST becomes a tolerance test If V LAST VER DIST
119. ates Details The VSHOW MODEL instruction may be used to display and make queries about one or all correlation templates in vision memory The vision system displays a specific template by copying the template from system memory into the frame store currently displayed The template is centered in the frame store If all templates are specified for display a list of the identifying numbers of all known templates is displayed The VSHOW MODEL instruction may also be used to display and make queries about one or all OCR fonts in vision memory Fonts are displayed in the Vision display window If all fonts are specified a list of the identifying numbers of all known fonts is displayed If a specific font is specified the characters in the font are displayed along with the following information e The approximate height of the font as specified in the VDEF FONT instruction e The color of the font characters black on white or white on black Adept Vision Reference Guide Rev A 235 VSHOW MODEL Program Instruction e The base orientation This angle is set when the font is first trained see the instruction VTRAIN MODEL When the font is subsequently trained or recognition verification is performed see VOCR the features in the OCR window are rotated as necessary to match the base orientation e For each character in the font the vision system displays the number of times that the character has been trained The above information
120. ays the corresponding single grip position This parameter is ignored if no display mode is requested edge_num Optional real valued expression that indicates if the model edge numbers are to be shown The default is 0 which means the numbers are not shown All edge numbers are shown if edge_num is 1 If edge_num is a positive number the number of that edge is shown AdeptVision Reference Guide Rev A 225 VSHOW Program Instruction Details 226 The type of model being displayed is determined by the name which has been trained as either a prototype or an ObjectFinder model The name is checked against the lists of prototypes first for backward compatibility Therefore you should not use the same name for a prototype model and an ObjectFinder model The VSHOW instruction may be used to display and acquire information about a single model After each use of the VSHOW instruction the VFEATURE real valued function may be used to obtain information about the model VSHOW is similar to the VLOCATE program instruction except that VSHOW works with pretaught models and VLOCATE works with objects found by VPICTURE or VWWINDOW VSHOW optionally displays the specified model Thus with display disabled you can use the instruction to get the names of the loaded models one at a time without disturbing the display The mode bits are defined as follows they all default to zero Bit 1 LSB Show model location constraints
121. b mode 2 below by simply VSELECTing the frame Also if the frame being acquired into is different from the frame store currently selected a future frame grab is automatically treated as a quick frame grab This distinction is important because almost all of the vision instructions operate on images acquired via quick frame grabs VPICTURE in mode 0 processes the pregrabbed image If no pregrabbed image is being held the previous image is reprocessed This is useful for experimenting with different switch and parameter settings Note however that the settings of the gain offset and threshold parameters do not have any effect upon reprocessing of a picture The VEDGE and VTHRESHOLD instructions may be used to compute new edge or binary images from an existing grayscale image If there is no pregrabbed image mode 0 is identical to VWINDOW cam 1 dmode how_many 1000 Adept Vision Reference Guide Rev A 197 VPICTURE Monitor Command and Program Instruction 198 Mode 2 is provided as a quick frame grab mode for measurement and inspection instructions such as VRULERI and VFIND LINE for image processing instructions such as VHISTOGRAM and VSUBTRACT and for the VWWINDOW instruction VPICTURE in mode 2 only grabs an image performing no automatic region or boundary analysis Mode 2 takes from 16 50 milliseconds to complete if the V BINARY system switch is enabled By comparison VPICTURE operations in modes
122. below where a score of 100 indicates complete ambiguity or perfect match if in the diagonal where the character is compared to its own model Color Score Dark gray 0 to 10 Medium gray 10 to 20 Dark green 20 to 30 Green 30 to 40 Yellow 40 to 50 Orange 50 to 60 Red 60 to 100 The vision system tries to make all the scores outside the diagonal below 30 When a score cannot be reduced more for some reason a small black plus sign is drawn in the patch of color Adept Vision Reference Guide Rev A 279 VTRAIN MODEL Program Instruction When a font is stored to disk via VSTORE the character information accumulated as a result of training is stored but the results of planning are not stored The data structures created during planning are very large Thus rather than storing and loading large disk files which in itself takes considerable time smaller files are used Consequently after a font is VLOADed it must be planned before use A 26 character font requires approximately 35 Kb of memory Example This routine is a simple interface for training on a subset of the characters in a font A window is displayed in the Vision display window overlaying live video so you can position the text inside it After you press Enter a picture is taken and the characters in the window are processed To stop training type any character before pressing the Enter key PROGRAM train Sstr Al already been defined B
123. best correlation score in the range 0 to 1 where 1 is perfection Elements data i 1 and data i 2 receive the X and Y coordinates in millimeters of the center of the template where it received the best correlation Location accuracy is only to one pixel Optional array index that identifies the first element to be defined in the array data The default is 0 If a multiple dimension array is specified only the right most index is incremented as the values are assigned Optional variable that returns the actual depth of a hierarchical search Searches will not go to the deepest level specified in max_depth if the template becomes too small the search area size is too close to the template size or the template degrades as a result of reducing Real valued expression specifying the correlation template number in the range 1 to 99 inclusive Optional integer value specifying the image buffer region to search for a match with a correlation template tp1num Image buffer regions specify both a size and a frame store see the description of VDEF AOI The template size is reduced if necessary to a multiple of four pixels If an image buffer region is not specified the currently selected frame store is used The template size is reduced if necessary to a multiple of four pixels If an br is not specified the currently selected frame store is used Not currently used Real valued expressions specifying the p
124. binary or edge image the image you see when in VDISPLAY mode 2 The image data is binary for type 2 VGETPIC Therefore for efficiency the data is packed eight pixels to a string character with the left most pixel in the least significant bit No partial characters are packed so x0 and dx are automatically adjusted to have the image window read start and stop on an eight pixel boundary ax ends up being a multiple of eight The actual starting coordinates and dimensions used are stored in the header string NOTE The only required parameter is pic All the other parameters default so that the entire image is saved The sample rate s_rate and sample mode s_mode parameters are supported for the image frame stores When the sample rate is greater than one the image stored in pic 1 is significantly smaller The sample mode determines the method of image reduction For example if the sample rate is 2 the image is shrunk by a factor of 2 in both the X and Y directions If the sample mode is 0 the image is reduced simply by using every other pixel on every other line If the AdeptVision Reference Guide Rev A Program Instruction VGETPIC sample mode is 1 each pixel in the reduced image is the average of the pixels in a 2x2 neighborhood Likewise if the sample rate is 6 mode 0 would use every 6th pixel on every 6th line and mode 1 would determine each pixel of the reduced image to be the average o
125. c circle Coordinates and radii are in millimeters For line edges elements data i 6 through data i 8 are not defined Optional integer value that identifies the first array element to be defined in data Zero is assumed if the index is omitted If a multiple dimension array is specified only the right most index is incremented as the values are assigned String expression that specifies the name of a prototype a subproto type or If a subprototype is specified the string must have the form namel name2 where namel is the name of a prototype and name2 is the name of the prototype s subprototype If edge information about a region in the image that is not a prototype is being requested this parameter should not be specified defaulted or it should be Adept Vision Reference Guide Rev A 87 VEDGE INFO Program Instruction edge_num Real valued expression that specifies the number of the edge of the region prototype or subprototype for which information is requested Details Knowing the positions and lengths of prototype edges could be useful in a V program that automatically places VFIND LINE and VFIND ARC operators on selected edges of prototypes to determine edge positions with high accuracy The operators would be applied after prototype recognition The instruction VEDGE INFO retrieves information about the edges and indirectly the corners of a prototype The edg
126. camera images may be VPUTPIC d back into the image memory and reprocessed To restore an old image for reprocessing a quick frame grab VPICTURE in mode 2 should be performed first Then the saved grayscale and binary image may be restored via a type 1 VPUTPIC or a binary image may be restored via a type 2 VPUTPIC The VWINDOW instruction or a VPICTURE in mode 0 repicture processes the data in the binary image The VWWINDOWI and ruler instructions use either the binary or the grayscale image depending on the type of operation If all the optional parameters are defaulted all the pertinent information is taken from the header string However a VPUTPIC of a saved grayscale image places the image in the frame store currently selected See the VSELECT program instruction See the description of the VGETPIC instruction for detailed information on the header string Example Reduce the full grayscale image by a factor of 4 in both the X and Y directions and then put it back zoomed up by the same amount Use averaging mode 1 when reducing the image VGETPIC 4 1 Spicture VPUTPIC 44 4 picturel Restore the same image but at its reduced size The image is repeated 16 times to fill the Vision display window FOR y0 1 TO 480 STEP 480 4 FOR x0 1 TO 640 STEP 640 4 VPUTPIC 1 Spicture x0 y0 END END Related Keywords VDISPLAY monitor command and program instruction VGETPIC prog
127. cameras are associated with physical cameras in blocks of four virtual cameras 1 4 are associated with physical cameras 1 4 virtual cameras 5 8 are associated with physical cameras 5 8 etc For each virtual camera one millimeter is made equal to one pixel and perspective calibration is not used This makes the system usable for setup or experimentation For precision work however a calibration procedure should be performed AdeptVision Reference Guide Rev A 139 VGETCAL Program Instruction Since the calibration arrays are normally filled by an Adept V utility program the programmer should not have to be completely familiar with the individual array elements However the contents of the arrays are listed in this manual in the description of the VPUTCAL instruction Exa mple VGETCAL 3 cal mp pm Get cam calibration data Related Keyword VPUTCAL program instruction 140 AdeptVision Reference Guide Rev A Program Instruction VGETPIC Syntax VGETPIC cam type s_rate s_mode pic r c shape x0 y0 dx dy Function Read all or part of an image into a string array Usage Considerations The VISION switch must be enabled the vision processor must be idle and vision model training must not be active for this instruction to be executed Parameter cam Optional real valued expression that specifies a virtual camera number The default is 1 This parameter is currently ignored type Optional real valued expre
128. cases the value should be at least 3 However a value of 2 may be sufficient for some applications AdeptVision Reference Guide Rev A Real Valued Function VFEATURE ITEM 27 The number of features in the finder model VSHOWed All of the line segments and circular arcs in the model are counted even if the weight after multiple instance training is zero ITEM 28 Maximum pixel variation measured in units of pixels This parameter controls the fitting of features to edges ITEM 29 Maximum location distance measured in units of pixels This parameter controls the distance between proposals It is used in a strategy that attempts to prevent too many proposals from being made in the same location Example F VFEATURE 1 THEN TYPE Recognition time VFEATURE 18 END Related Keywords VLOCATE program instruction VSHOW program instruction Adept Vision Reference Guide Rev A 105 VFIND ARC Syntax VFIND ARC VFIND ARC Function cam mode dmode effort type data i Program Instruction ibr cam mode dmode effort type data i 1 xc yc Y rr ang0 angn Fit a circular arc to an image edge bounded by a window that is shaped like a ring or a ring segment Usage Considerations If the calibration being used includes correction for perspective distortion the correction is applied to the given center of the arc xc yc and to the returned values describing the fit arc A
129. ch Adept Vision Reference Guide Rev A 113 VFINDER Program Instruction Syntax VFINDER cam type dmode how_many_total times ibr Function This instruction performs ObjectFinder recognition using the planning associated with the given virtual camera Parameter cam Real valued expression indicating the virtual camera s planning to use and parameters to read type Type of recognition operation to perform Default value is 1 1 ObjectFinder recognition 2 ObjectFinder recognition for multi instance training dmode Real valued expression that specifies the display mode to use when displaying the border of the window 1 No draw Erase Draw solid default Complement Draw dashed Complement dashed AeUNBeO how_many_total Specifies the total number of objects to find In the case of multiple models planned together this is regardless of which models are found DEFAULT 1 find as many objects as possible If 0 is given none will be found but the image is still processed through feature extraction times Array of max_time information for the recognition process The whole array is optional defaults are as if times 0 0 0 How many max_time parameters to follow 1 Max time to first part defaults to V MAX TIME cam 2 Max time to subsequent parts defaults to times 1 ibr Integer value specifying the image buffer region within which to search The AOI must be a no
130. ch font is completely independent A font may contain up to 94 characters including symbols such as and All the characters must be in the ASCII range 33 to 126 In particular the space character ASCII 32 is not permitted The characters defined to be in a font are actually just labels The vision system has no built in strategies that are character specific Normally you would use A to label the character A However OCR performance will not be degraded if you use to label the character A and train the system as such Then when the vision system subsequently sees an A character it will report that it saw a Wy See the AdeptVision User s Guide for an overview of the AdeptVision OCR capability Example The following instruction defines font 1 as a black set of numerals approximately 24 pixels tall VDEF FONT 0 1 0123456789 24 TRUE Related Keywords VDELETE monitor command and program instruction VOCR program instruction VSHOW MODEL program instruction VTRAIN MODEL program instruction AdeptVision Reference Guide Rev A 63 VDEFGRIP Program Instruction Syntax VDEFGRIP proto grip mode num_fngrs trans i w j h k VDEFGRIP proto 0 Function Define the shape and position of a robot gripper for clear grip tests Usage Considerations The VISION switch must be enabled and the vision processor must be idle for this ins
131. ch look like little knots The edges are displayed in a unique color to distinguish them from other graphics Example 230 Show the results of subsequent edge fitting ENABLE V SHOW EDGES AdeptVision Reference Guide Rev A System Switch V SHOW EDGES Related Keywords VFIND ARC program instruction VFIND LINE program instruction VFIND POINT program instruction VPICTURE monitor command and program instruction VWINDOW program instruction AdeptVision Reference Guide Rev A 231 V SHOW FEATS System Switch Syntax V SHOW FEATS camera Function Enable the special display of features used for ObjectFinder recognition Usage Considerations A change to this switch takes effect when the next region is displayed in the Vision display window with a graphical display mode in effect This is an array of switches one for each virtual camera See the AdeptVision User s Guide for details on setting switches Details When V SHOW FEATS is enabled the features used for ObjectFinder recognition are displayed in the Vision display window overlaying the displayed regions Each connected pair of lines or arcs is joined at a corner which looks like a little knot The bounds are displayed in a unique color to distinguish them from other graphics If VSHOW FEATS is disabled the system does not perform arc and line fitting for ObjectFinder recognition since it is not needed For applications that do not use recognition disab
132. ciated with the selected frame store By default the previous association with the frame store is unchanged if camera is omitted Details There are 32 maximum virtual grayscale frame buffers The VSELECT instruction selects the virtual frame buffer to access Operations such as VRULERI VWWINDOWL VFIND LINE VHISTOGRAM and VGETPIC access the currently selected frame buffer By default the current frame buffer is the one selected by a VPICTURE operation Using VPICTURE an image can be acquired into any frame buffer The VSELECT instruction performs the same function as the sel_ibr parameter to VPICTURE Since VSELECT is a separate instruction however a program can acquire two camera images into different frame buffers and then use VSELECT to quickly switch back and forth between them while performing analyses 1 The actual number is configurable with the DEVICE program instruction see the Adept Vision User s Guide 220 Adept Vision Reference Guide Rev A Program Instruction VSELECT When a frame buffer is selected the virtual camera number used with the VPICTURE operation that acquired that image is automatically selected along with the calibration data and all the system switches and parameters associated with that camera However if a camera parameter is supplied in the VSELECT instruction the calibration data switches and parameters for that camera are selected This capability is useful wi
133. ck All pixels with an intensity less than or equal to this threshold are set to black and pixels with intensities greater than this value are set to white The V 2ND THRESH system parameter provides a dual threshold capability See the description of V 2ND THRESH for more information V THRESHOLD and V 2ND THRESH are for use during binary image processing For grayscale processing there is an edge threshold parameter called V EDGE STRENGTH The correct value for V THRESHOLD depends on the particular application The VAUTOTHR program instruction may be used in most cases to automatically determine the best threshold V THRESHOLD must be assigned an integer value in the range 0 to 127 inclusive It has the initial value 63 Example Set all grayscale pixel values above 100 to white and all others to black PARAMETER V THRESHOLD 100 Related Keywords VAUTOTHR monitor command and program instruction V 2ND THRESH system parameter V BINARY system parameter 262 Adept Vision Reference Guide Rev A System Switch V TOUCHING Syntax V TOUCHING camera Function Determine whether or not objects may be touching in the image Usage Considerations This switch is for use only with prototype recognition V RECOGNITION must be enabled for this switch to have any effect A change to this switch takes effect when the next VPICTURE command or instruction or VWINDOW instruction is executed V TOUCHING is assumed to be enabled if the
134. ct_pt 1 contain the X Y location of the object centroid with correction for perspective distortion AdeptVision Reference Guide Rev A Upgrading Program Code Introduction Compatibility Summary Adept Vision Reference Guide Rev A 313 Appendix D Introduction Intoduction Program code written for systems prior to version 11 0 needs to be upgraded All version 11 0 and later compatible program code will continue to work with version 12 1 systems The following section provides a compatability summary for users that need to upgrade pre 11 0 program code This is not a summary of all changes that occurred prior to version 11 0 It is a summary of operations that have changed and may require modifications to upgrade pre 11 0 program code so it will execute correctly in version 11 0 and later systems Compatbility Summary 314 Maximum value for V LAST LINE is now 480 instead of 484 The system parameters V FIRST COL V LAST COL V FIRST LINE and V LAST LINE are no longer used as a clipping window for VOCR V FIRST COL etc no longer define the AOI for the following instructions VADD VAUTOTHR VCOPY VCONVOLVE VEDGE VHISTOGRAM VMORPH VSUBTRACT VTHRESHOLD The old frame numbers 1 and 2 will no longer be accepted 1nnn e g 1001 and 1002 must be used When a frame number of 1nnn e g 1001 is used the entire frame is processed as before The default value of V OFFSET is set to 255 instead of 127 when the
135. ction VDEELUT Syntax VDEF LUT lut num lut Function Define a grayscale and binary look up table for mapping graylevel and binary values during a VCOPY operation Usage Considerations Available only with the AdeptVision Enhanced VXL Interface option Parameters lut num Integer from 1 to 133 maximum that identifies the 1ut to define lut Array of integer values in the range 0 255 that define the pixel mapping Details The maximum 1ut_num is configurable using the DEVICE command The default value is 3 The maximum is 133 LUTs taking 33 Kb The 1ut 1 is an array of 256 values index in range 0 to 255 where each value must range from 0 to 255 The indices and the values in the array are both pixel values A pixel value is an 8 bit number Bits 1 thru 7 are the graylevel value 0 127 and the binary bit is bit 8 Therefore indices 0 127 to the array will transform the graylevels for which the binary bit is 0 and indices 128 255 will transform the graylevels for which the binary bit is 1 In other words this is an 8 bits in 8 bits out LUT The one exception is when the VCOPY is used to subsample or zoom Then the LUT is only 7 bits in That is only the indices 0 127 are used and they are applied AFTER the subsample or zoom operation In this case the LUT is still 8 bits out Exa mple Define a look up table that turns all graylevel values 50 or lower to blac
136. d Keyword Description V MAX TIME Set the maximum time allowed for the vision system to analyze a region during prototype recognition or OCR V MAX VER DIST Set the pixel tolerance for determining boundary coincidence during the verification of prototype to image matches V MIN AREA Set the minimum area below which the vision system ignores regions V MIN HOLE AREA Set the minimum area below which the vision system ignores holes V MIN LEN Set the minimum length of features to be used for feature pairs V MIN MAX RADIT Enable the feature that for each region in the image finds the two points on the perimeter that are closest to and farthest from the region centroid VMORPH Perform a morphological transform on a binary image frame VMORPH Set the offset for the incoming video signal that is program the zero reference for the A D converter VOCR Perform Optical Character Recognition OCR or text verification in a rectangular image window V OVERLAPPING Determine whether or not objects may be overlapping in the image and still be recognized V PERIMETER Enable computation of the lengths of region perimeters VPICTURE Grab an image for processing VPLAN FINDER Set up the type of planning used by the Finder when locating models VPUTCAL Give the system arrays filled with vision calibration data VPUTPIC Store into a frame store an image saved previously w
137. d The effects of changes to this parameter are visible in the Vision display window only in a special display mode Holes with areas smaller than V MIN HOLE AREA will still appear in display modes 1 and 2 This is an array of parameters one for each virtual camera See the AdeptVision User s Guide for details on setting parameters Details Holes that contain fewer than V MIN HOLE AREA number of pixels are ignored by the vision system during object recognition clear grip testing and standard ruler measurements This parameter should be set large enough so that the vision system does not waste time analyzing noise spots and small enough so that no region of interest is discarded This parameter must be assigned an integer value in the range 1 to the value of the parameter V MIN AREA inclusive The system parameter V MIN HOLE AREA is set to 8 when the V and AdeptVision systems are loaded into memory from disk Example Ignore holes that are smaller than 10 pixels in area PARAMETER V MIN HOLE AREA 10 Related Keywords V MAX AREA system parameter V MIN AREA system parameter 176 AdeptVision Reference Guide Rev A System Parameter V MIN LEN Syntax V MIN LEN camera Function Set the minimum length of features to be used for feature pairs Usage Considerations This parameter is used by VTRAIN FINDER and VPLAN FINDER See those program instructions for details This is an array of parameters one for e
138. d to determine which pixels are black and white Therefore a different binary threshold can be used for each finder Raw binary line finders type 1 use the raw binary frame store This is the data shown in VDISPLAY mode 2 This is binary or edge data depending on the setting of the V BINARY system switch Run length line finders type 0 use the processed binary data produced by VWINDOW or VPICTURE in modes 1 or 0 This data is shown in VDISPLAY mode 3 Note that system parameters such as V MIN AREA and V FIRST COL are used during picture processing and therefore will affect this type of finder whereas they will not be taken into account for the other binary finders type 1 and 2 Again the image data is binary or edge data depending on the setting of the V BINARY system switch Fine edge line finders type 2 are the highest precision finders The edge points used in the least squares fit algorithm are found in the grayscale frame store with subpixel accuracy The image data used is visible in the Vision display window in VDISPLAY mode 1 At the center of the VFIND LINE search window is the guideline the user s estimate of the edge location The guideline is defined by a point the center point and an angle The search area is further specified by the length of the guideline breadth of scan the width of the search window range about the guideline within which the vision system is to search and an
139. d the sum of the negative ones should be close to 128 and 128 respectively These guidelines will prevent overflow when computing the new center pixel value If the contrast in the image is low however larger terms may be used If a convolution is defined that is not centered in the 7x7 array it will shift the resulting image To compensate for the shift you should recalibrate the vision system with that convolution enabled If an edge convolution is used the V BINARY system switch should be enabled Otherwise the vision system will be finding edges of the edges When you use a convolution the image boundaries as defined by VDEF AOI should be moved into the image a pixel or two to mask out the noise generated when the convolution straddles the physical image boundary 1 1 1 2 1 3 1 4 1 5 1 6 1 7 2 1 2 2 2 3 2 4 2 5 2 6 2 7 3 1 3 2 3 3 3 4 3 5 3 6 3 7 4 1 4 2 4 3 4 4 4 5 4 6 4 7 5 1 5 2 5 3 5 4 5 5 5 6 5 7 6 1 6 2 6 3 6 4 6 5 6 6 6 7 7 1 7 2 7 3 7 4 7 5 7 6 7 7 Figure 2 4 Arrangement of Elements of Convolution Matrix AdeptVision Reference Guide Rev A Program Instruction VDEECONVOLVE NOTE Only the innermost 3x3 elements are considered with standard vision systems Example The following sequence of program instructions creates a new 3x3 low pass filter convolution that is ide
140. d to extract lines and arcs The lines and arcs are paired up in various ways to form intrinsic feature pairs These can be later used to match with similar pairs formed from recognition images to make proposals of instances of the model These lines arcs and pairs define the geometric portion of the model There are three system parameters described below that control the processing 1 The system parameter V EDGE STRENGTH vc controls the threshold for edge points extracted during preprocessing 2 The system parameter V MAX PIXEL VAR vc controls the primary fitting of lines and arcs A secondary operation performs a least squares fitting to the edge points which greatly improves the accuracy 3 The system parameter V MIN LENT vc restricts the pairs by setting a minimum length for the features allowed to be used to make pairs The system switch V FIT ARCS vc should normally be enabled but can be disabled to some advantage in certain trouble cases If an object is mostly lines but contains some segments that occasionally fit as arcs due to noise in the image you can force the system to fit only lines and avoid the additional computation time required to fit arcs Input parameters using virtual cameras All switches and parameters should be at the default settings except as noted here The following settings are required PARAMETER V MIN AREA vc 4 PARAMETER V MIN HOLE AREA vc 4 AdeptVision Reference Guide Rev A
141. d virtual camera and fills the queue with the results of the new image An object must be VLOCATEd in order to obtain subprototype information see the VSUBPROTO instruction or gap information see the VGAPS instruction 1 VLOCATE retrieves one object from the queue After each VLOCATE the VFEATURE real valued function may be used to obtain information about the found object 160 AdeptVision Reference Guide Rev A Program Instruction VLOCATE NOTE Subprototype information and gap information does not apply to the ObjectFinder tool The mode bits are defined as follows they all default to zero Bit 1 LSB Wait 0 versus No wait 1 mask value 1 Bit 2 Bit 3 Wait 0 The VLOCATE instruction will stop program execution until any region is located or a specific prototype is recognized depending on the find any find particular bit see below Once the VLOCATE is satisfied control returns to the V program even if image processing is not yet complete that is there are more regions to process If the VLOCATE is not satisfied program execution continues when the entire image has been processed No wait 1 Program execution will continue without waiting Default Wait mode 0 Find any 0 versus Find particular 1 mask value 2 Find any 0 Any object located for the given virtual camera will satisfy the VLOCATE the next one in the queue is returned in the desired order The name para
142. dept recommends that you use the first syntax Parameters cam mode dmode effort type data 106 Optional real valued expression that specifies a virtual camera number The default is 1 The camera number is used to determine which V EDGE STRENGTH parameter to use Optional bit field expression The default is 0 In summary the bit mask values and their meanings are listed here see below for details Circle color 0 dark 1 light Find 0 center only 2 radius only 4 both Search start position O center 8 inner 16 outer Optional real valued expression specifying the display mode for this operator The choices are 1 no draw 0 erase 1 draw solid 2 complement 3 draw dashed and 4 complement dashed The default is 1 draw solid Optional real valued expression indicating the effort level to use when finding the arc The effort level ranges from 1 least effort to 100 maximum effort The default is 20 Optional real valued expression specifying the type of arc finder 2 dynamic binary 1 raw binary 0 run length binary 2 fine edge Default is 2 fine edge Real array describing the outcome of the arc fit data i 0 TRUE if an arc was fit otherwise FALSE AdeptVision Reference Guide Rev A ibr xc ye rr ang0 angn Program Instruction VFIND ARC data i 1 TRUE if any part of the search window falls off the Vision display window data i 2 dat
143. deptVision VXL option and V operating system version 12 1 or later See Appendix D for details on using existing program code with 12 1 systems Conventions Many operating system options require you to hold down the keys marked Shift or Ctrl and then press another key These types of key combinations are shown as Ctrl C which means you should hold down the Ctrl key and press the C key Related Publications This reference guide details the extensions added to the V language with the AdeptVision VXL option It is a companion to the V Language Reference Guide and V Operating System Reference Guide which cover the basic V language and operating system The following manuals may be required by your system Table 1 1 Related Publications Manual Material Covered AdeptVision User s Guide Enhancements to the V operating system that are added when the AdeptVision option is installed Instructions for Adept Utility Programs Instructions for running various setup and configuration software utilities Adept MV Controller User s Guide Instructions for setting up configuring and maintaining the controller that runs Vt Robot motion or force device user s guides if connected to your system Instructions for installing and maintaining the motion device connected to your system Manual Control Pendant User s Guide if connected to your system Using the manual co
144. distortion is present Using the 3 by 3 transformations instead of just the scale transformation takes more processing time so their use is optional The vision system uses only the scale transformation if the G and H elements in the pmm to pix transformation are zero When shown a square calibration target the Adept calibration program always computes the perspective transformation and then derives the scale transformation from it The small translation component in the perspective transformation aligns the perspective and scale transformations in the center of the screen The small rotation component aligns the 45 degree lines in millimeter space and pixel space Based on the pmm to pix and pix to pmml transformations the LOADAREA program derives two additional transformations pmm to mm and mm to pmm These are provided to allow V application programs to correct for perspective distortion when the vision system does not do so In summary the four transformations are Adept Vision Reference Guide Rev A 309 Appendix C 310 pmm to pix pix to pmm pmm to mmf mm to pmm Calibration Arrays Perspective millimeters to raw pixels The vision system uses this transformation to convert user specified millimeter coordinates such as the start of a ruler to pixels This is the inverse of pmm to pix The vision system applies this transformation wh
145. e 8 Shape 9 dims dim4 Shape 10 ang1 dim4 dim3 ang2 Figure 2 3 Shape Parameters for Arc Shaped Tools Adept Vision Reference Guide Rev A 53 VDEFAOI Program Instruction Before an AOI can be used by a vision tool it must be combined with a virtual frame buffer number The VDEF AOI instruction specifies only the area of interest information You must add a frame buffer specification to the area of interest number to create a complete image buffer region An image buffer region is a 4 to 6 digit value VDEF AOI specifies the left 1 to 3 digits and ignores the right three digits The frame store information must be added to the right three digits to complete the image buffer region specification This allows the same area of interest to be used in different virtual frame buffers See the example below There are three physical frame stores in an AdeptVision VXL system Two are for vision operations and one is for displaying vision data The two physical frame buffers used for vision operations are further divided into virtual frame stores Virtual frame buffers are allocated with the V DEVICE command See the V Language Reference Guide for details on DEVICE See the AdeptVision User s Guide for a summary of using DEVICE to configure virtual frame stores AOIs that extend beyond the virtual frame buffer are clipped The form of a complete image buffer region is NNN VV P area of interest
146. e of V BACKLIGHT AdeptVision Reference Guide Rev A 31 V BACKLIGHT Example System Switch Enable processing of backlit objects for virtual camera 3 ENABLE V BACKL GHT 3 Get the current setting for virtual camera 1 val SWITCH V BACKLIGHT 1 Related Keywords V BINARY system switch VWINDOW program instruction VWINDOWB program instruction VWINDOWI program instruction 32 AdeptVision Reference Guide Rev A System Switch V BINARY Syntax V BINARY camera Function Enable or disable automatic edge image generation at VPICTURE time Usage Considerations This switch may become obsolete in future versions Its functionality is identical to performing a VEDGE instruction after a VPICTURE 1 2 instruction A change to this switch takes effect when the next VPICTURE command or instruction is executed This is an array of switches one for each virtual camera See the general description of switches in AdeptVision User s Guide for syntax details Details Normally this switch is enabled If disabled it will affect the operation of VPICTURE modes 1 1 and 2 as described below For VPICTURE modes 1 and 2 it will start a VEDGE operation immediately following the completed acquisition into the virtual frame buffer For VPICTURE mode 1 a VEDGE operation is performed prior to processing of the image In this case the VPICTURE instruction will not complete until a
147. e a portion subname is the name of the subprototype being defined For the monitor command these must be specified by an unquoted string see the example below For the program instruction they must be specified by a string constant variable or iy expression containing the separator see the example below first_edge Real valued expression that specifies the edge number of the prototype that is to be the first edge of the subprototype last_edge Real valued expression that specifies the edge number of the prototype that is to be the last edge of the subprototype Note that last_edge may be less than first_edge Details This instruction defines a subprototype of the prototype proto to be the series of edges from first_edge to last_edge The subprototype must consist of edges from only one region of the subprototype Thus for example two holes cannot be grouped together to form a single subprototype If first_edge equals last_edge the subprototype consists of a single edge Once the subprototype is defined it has its own edge numbering Thus edge first_edge of proto is edge 1 of the subprototype A subprototype remains associated with a prototype until the subprototype or the prototype is VDELETEd A subprototype is automatically saved and restored when its associated prototype is VSTOREd and VLOADed Subprototypes may also be referenced by VEDGE INFO VG
148. e angle formed by the two edges is associated with each corner class Also acceptable minimum and maximum lengths are associated with each line edge of a feature class For arcs there are acceptable minimum and maximum angular ranges minimum and maximum radii and a convex or concave indication Each corner and edge in the image is compared with all the feature classes When the types correspond and the measures of the corner or edge fall within the acceptable ranges of the class the corner or edge is placed on the association list for the feature class Some corners and edges in the image may not be associated with any classes whereas others may be associated with multiple classes Propose Prototype to image Matc hes 304 This step proposes that prototypes are present in the image at specific positions and orientations These prototype to image match proposals are based on the feature classifications made in the previous step During training the feature classes are ordered according to uniqueness and clarity At run time in this Adept Vision Reference Guide Rev A Appendix B Verify Match order each class is considered until all image edges have been accounted for or until all possible matches have been proposed Each image feature and prototype feature associated with a class is a possible match However such matches must be confirmed before an official proposal can be made and step 7 begins Confirmation is a partial ver
149. e for this service except for any long distance toll charges Simply call either number and follow the instructions to have information faxed directly to you Adepton Demand Web Page If you have access to the Internet you can view Adept s web page at the following address http www adept com The web site contains sales customer service and technical support information 16 Adept Vision Reference Guide Rev A Desc nptions of Vision Keywords This chapter presents detailed descriptions of all the V system keywords related to vision These keywords are available only to systems with the AdeptVision VXL option The global system switch VISION enables and disables all vision processing If this switch is DISABLEd the system will behave as if the vision option is not installed The VISION switch default value is ENABLEd The vision system includes the following types of keywords e Program instructions e Functions e Monitor commands e System parameters e System switches The keywords are presented in alphabetical order with the description for each keyword starting on a new page Note that some keywords may belong to more than one keyword type The keyword is shown at the top of each page The sections described on pages 18 and 19 are included for each keyword as appropriate AdeptVision Reference Guide Rev A 17 KEYWORD Keyword Type Syntax This section presents the syntax of the keyword The keyword is shown
150. e is either very rough or some extraneous edges were detected and used in the arc fitting If edge filtering is enabled data i 7 and data i 8 more specifically indicate the most distant edge points inside and outside the arc respectively Data item data it 6 is the maximum of data i 7 and data i 8 Accuracy is more dependent on the angular range of the arc than on the effort level Angular ranges of 180 degrees or more provide very accurate subpixel results Fitting an arc with an angular range of less than 45 degrees can have very large errors This assumes that both the center and radius are being computed The results are much more accurate if one or the other is known in advance AdeptVision Reference Guide Rev A Program Instruction VFIND ARC Example The following instruction will use a fine edge default type arc finder to look for a light circle on a dark background mode bit 1 set computing both the center and radius mode bit 3 set The effort level is 50 The estimated center location and radius are 30 46 and 10 respectively The search range is 5 the angular range is 360 degrees 0 to 0 VDEF IAO 2001 5 30 46 10 5 VFIND ARC 1 BOR 4 50 data 2001 Related Keywords VDEFAOI program instruction VFIND LINE program instruction VFIND POINT program instruction V MAX PIXEL VAR system parameter V MAX SD system parameter V SHOW EDGES system swit
151. e loaded into memory from disk AdeptVision Reference Guide Rev A System Parameter V 2ND THRESH Examples Set dual thresholds for a region of interest with intensity range 45 to 67 PARAMETER V 2ND THRESH 45 PARAMETER V THRESHOLD 67 Related Keywords VAUTOTHR monitor command and program instruction V BINARY system parameter V THRESHOLD system parameter AdeptVision Reference Guide Rev A 23 VABORT Monitor Command and Program Instruction Syntax VABORT task_id Function Abort any active vision processing and all pending vision operations associated with the given task number Para meter task_id Optional real valued expression specifying the number of a V program execution task The special task number 2 indicates all tasks The special task number 1 indicates self meaning the task issuing the VABORT Otherwise task_id is an integer value in the range 0 to N where N is the largest task number supported by your V system The default value is 1 meaning self Details Most users will never need to use VABORT When a user aborts a V program that is executing a vision instruction or when the user aborts a vision monitor command by typing Ctrl C V automatically aborts the associated vision processing In fact V automatically sends the vision processor a VABORT request for the respective execution task whenever a task is primed for execution starts executing or completes executio
152. e rows read and stored 25 chars_per_row Number of characters per row used in pic 27 clipped Boolean TRUE if clipped in either direction AdeptVision Reference Guide Rev A 143 VGETPIC Program Instruction Example VGETPIC Ssavpic Save the current full image VPUTPIC Ssavpic Restore the image saved above See the description of VPUTPIC for more examples Related Keywords VDISPLAY monitor command and program instruction VPUTPIC program instruction VSELECT program instruction 144 AdeptVision Reference Guide Rev A Program Instruction VGET TRANS Syntax VGET TRANS array i Function Return the value of the current vision transformation Usage Considerations Each task has its own vision transformation so VGET TRANS must be issued in the correct task Para meters array Variable identifying the array to receive the vision transformation i Optional integer expression specifying the starting array index The default is 0 Details The current vision transformation definition is returned in the array elements arrayli Always returns 1 array i 1 X offset of the vision transformation array i 2 Y offset of the vision transformation array i 3 Rotation of the vision transformation array i 4 Scale of the vision transformation Exa mple Rotate the current vision transformation by 90 VGET TRANS t_def VDEF TRANS t_def 1 t_def 2 t_def 3 90 Related Keyword VDE
153. e training use VFINDER with type 2 arg Indicates the value 1 Symmetry of model 2 Unused 3 Verify percent to assign to model value The value to assign to the designated parameter Mode 4 Feature weight setting also feature deletion arg Feature number 1 to num_features 1 means all features value New weight to assign O deletes the feature deletion means that the feature is not used 1 100 relative weighting for verification The feature is still used to form feature pairs Mode 6 Display a model feature arg Number of the feature to display Mode 7 Display a model feature in RED arg Number of the feature to display AdeptVision Reference Guide Rev A Program Instruction VTRAIN FINDER Mode 8 Accept the last trained additional instance arg Low weight cutoff used for multi instance training and is the threshold on feature weights on a scale of 0 100 feature weights below the low weight cutoff are set to zero Mode 9 Do not accept the last trained additional instance arg Low weight cutoff Mode 10 Clear all multi instance training statistics Leave the first instance arg Low weight cutoff Mode 11 Apply low weight filtering arg Low weight cutoff Mode 12 Undo the last accepted instance arg Low weight cutoff Details When normal initial training is performed a new model is created and assigned the given input values Additionally the ibr is processed to create edge pixels and these are further processe
154. ea to be detected A B C OO O V SUBTRACT HOLE enabled VSUBTRACT HOLE disabled V SUBTRACT HOLE disabled Total area edge region only Total area edge region hole Break causes hole area to gray area is excluded gray area is included merge with background Total area is significantly reduced This switch does not affect the calculations enabled by the system switches V CENTROID V PERIMETER V MIN MAX RADIL and V2ND MOMENTS AdeptVision Reference Guide Rev A 255 V SUBTRACT HOLE System Switch Example Subtract hole areas when computing region areas ENABLE V SUBTRACT HOLE Related Keywords V MAX AREA system parameter V MIN HOLE AREA system parameter V MIN AREA system parameter 256 Adept Vision Reference Guide Rev A System Parameter V SYNC STROBE Syntax V SYNC STROBE camera Function Select synchronous or asynchronous firing of a strobe light when a picture is taken that is when a VPICTURE is executed Usage Considerations A change to this parameter takes effect when the next VPICTURE command or instruction is executed The V STROBE switch must be enabled in order for the VSYNC STROBE parameter to have any effect This is an array of parameters one for each virtual camera See the AdeptVision User s Guide for details on setting parameters Details The value of this parameter determines when a strobe light should be fired with respect to the camera read out cycle The choices are synchrono
155. earched at full resolution This process greatly increases the speed of correlation matching The system will not allow a search that results in a reduced image smaller than 8x8 Adept Vision Reference Guide Rev A 47 VCORRELATE Program Instruction pixels For example take an AIO that is being searched with mode 3 First the area of interest definition is reduced to 64x64 the template is reduced a corresponding amount and a search is made for a match Next a 128x128 area is searched in close proximity to the best match found in the 64x64 reduction Finally a search is made of the 256x256 area in close proximity to the best match found in the 128x128 reduction See Appendix F in the AdeptVision User s Guide for further details The speed of the correlation operation is independent of the complexity of the image The speed depends on the size of the template the size of the search window and the effectiveness of the hierarchical search Specifically if the template size is w by h pixels and the window size is W by H the time spent to correlate is proportional to f w h W w 1 H h 1 When the search window is the same size as the template the above expression is f w h and the operation runs fairly fast NOTE Correlation templates provide a way to save images to CPU memory or disk Copying a complete image to memory takes about 1 second Storing it to disk takes about 14 seconds Loading it from
156. ecified frame store This should pose no backward compatibility problems but it was not done in previous versions VMORPH VEDGE VTHRESHOLD VCONVOLVE VCOPY VADD and VSUB operations do not need to wait for the vision processor to be idle This should pose no compatibility problems and may allow some applications to run faster AdeptVision VXL has a third frame store for live video Therefore displaying live video will not erase a physical frame store This should pose no compatibility problems and may provide a useful addition to some applications Convolutions larger than 3x3 require the AdeptVision Enhanced VXL Interface option The OCR capability is no longer optional It is included with the basic system Adept Vision Reference Guide Rev A 315 Numerics 2nd moment data enabling collection of 20 A Adept address e mail 15 fax back service 16 web page 16 Adept MV Controller User s Guide 8 AdeptVision keyword summary 9 Adept Vision User s Guide 8 AIM 9 AOI and physical frame store 54 and virtual frame store 54 AOI see Area of interest Application questions 15 Applications Internet e mail address 15 Arc Finder Shape 111 Arc fitting enabling disabling 131 Area of interest defining 50 Arrangement of elements of Convolution Matrix 60 B Background differentiating from foreground 31 Binary images 9 10 12 13 22 26 28 29 33 69 70 80 85 146 167 168 180 181 183 198 207 253 260 262 280 2
157. ed in the text parameter are assumed to be in this order Each character in a font may be trained a different number of times Also a character may appear more than once in a training window The vision system ignores the spacing between characters Thus for example training on the characters Lot 34 would have the same effect as training on Lot34 All characters in the font should be trained 5 to 15 times for reliable operation The vision system does not allow a character to be trained more then 30 times When training the vision system samples of the text representing the range of acceptable appearance should be used If only a single sample of the text is available you should still show it to the vision system multiple times moving it around a little and rotating it slightly in both directions such as plus and minus 5 to 10 degrees to provide some variety in appearance Note that training on a font is different from prototype training With font training there is no system interaction with the keyboard or pointing device In particular there is no Are you sure confirmation Consequently you must make sure the window encompasses the given text before executing the VTRAIN MODEL instruction Otherwise you could corrupt the system s models of the characters The Example section below contains a simple V routine to help perform font training VTRAIN MODEL returns an error and does not modify the current charac
158. ed instead of the height This parameter is required if the operation op is 0 otherwise height is ignored black_chars Optional real valued expression that is interpreted as a boolean TRUE FALSE value A TRUE value indicates that the characters in the font are dark against a light background A FALSE value indicates the reverse situation This parameter is required if the operation op is 0 Otherwise black_chars is ignored Details This instruction defines an OCR font to the vision system It tells the system what characters are in the font the approximate size of the characters in the font and the relative color of the characters A font must be defined with this instruction before the vision system is trained on the font with the VTRAIN MODEL instruction 62 Adept Vision Reference Guide Rev A Program Instruction VDEEFONT If a font is already defined and VDEF FONT is performed with op equal to 0 the existing font is deleted and replaced with the new definition If op is 1 however an existing font is modified not deleted The only modification to a font that is allowed is to add or delete characters in the font The height and color of a font cannot be changed When a font is modified the given string chars defines the new set of characters Any training experience on characters that remain in the modified font is retained Up to 99 fonts may be defined memory permitting Ea
159. ed or is 0 VQUEUE returns the total number of prototype instances in all virtual camera queues If neither a camera nor a name is specified the total number of all objects in all queues is returned Examples Determine the total number of objects seen by all virtual cameras and the number seen by virtual camera 3 total objects VQUEUE object cnt 3 VQUEUE 3 Find out the total number of objects matching the prototype CASTING that have been seen by all virtual cameras and the number seen by virtual camera 7 casting cnt VQUEUE CASTING num castings VQUEUE 7 CASTING Related Keywords VLOCATE program instruction VQUEUE monitor command 210 AdeptVision Reference Guide Rev A System Switch V RECOGNITION Syntax V RECOGNITION camera Function Enable or disable prototype recognition by the vision system Usage Considerations A change to this switch takes effect when the next VPICTURE command or instruction or VWINDOW instruction is executed This is an array of switches one for each virtual camera See the AdeptVision User s Guide for details on setting switches Details This switch is normally enabled When V RECOGNITION is disabled the vision system operates as if no prototypes have been defined This is a useful operating mode when calibrating for example because then only the centroids of regions are needed If no prototypes are defined the state of V RECOGNITION has no e
160. ed than the one used at recognition time This ruler type is designed to detect edges that are close to perpendicular to the direction of the ruler If a ruler is far from perpendicular to an edge at the point where the ruler and edge intersect the ruler is less accurate and less sensitive to that edge A good rule of thumb is to keep the ruler within 20 degrees of perpendicular for best results The sensitivity of the ruler to edges is determined by the V EDGE STRENGTH parameter The higher the parameter value the stronger an edge must be in order to be detected The edges returned by a type 3 fine pitch ruler are similar to a type 2 ruler except the edge detection algorithm has been optimized for finding closely spaced edges For maximum accuracy this ruler type must be placed as perpendicular as possible to the edges being detected For straight rulers the distance to grayscale edges is accurate with varying degrees of subpixel accuracy depending mostly on ruler and edge orientation When both image edge and a linear ruler are exactly horizontal or vertical the distance to an edge has a standard deviation of about 0 04 pixel with a maximum error of 0 1 pixel For angled image edges the standard deviation is around 0 1 pixel with a maximum error of 0 3 pixel if the ruler is kept within 20 degrees of perpendicular to the edge The repeatability and absolute accuracy will depend on camera and lens quality as well as calibration and light
161. ees are most efficient Details The VFIND POINT instruction looks for edges in a search window It operates like a fat ruler looking along a wide path in search of an edge An example application would be to find the right most point on a circle when you know the circle s approximate position VFIND POINT could also be used to find the end of a pin VFIND POINT types 2 1 and 0 are binary point finders Binary edge points are black to white or white to black transitions The maximum accuracy is one pixel Dynamic binary point finders type 2 use the grayscale frame store This is the data visible in VDISPLAY mode 1 The values of the threshold parameters V THRESHOLD and V 2ND THRESH at the time the finder is executed are used to determine which pixels are black and white Therefore a different binary threshold can be used for each finder Adept Vision Reference Guide Rev A 125 VFIND POINT Program Instruction 126 Raw binary point finders type 1 use the raw binary frame store This is the data shown in VDISPLAY mode 2 This is binary or edge data depending on the setting of the V BINARY system switch Run length binary point finders type 0 use the processed binary data produced by VWINDOW or VPICTURE in modes 1 or 0 This data is shown in VDISPLAY mode 3 Note that system parameters such as V MIN AREA and V FIRST COL are used during picture processing and therefore will affect this ty
162. ement dashed The default is 1 draw solid dest_ibr Optional integer identifying the image buffer region that will receive the thresholded values ere abi Optional integer identifying the image buffer region that will be thresholded Details This instruction thresholds the frame store with the given threshold s This is useful after a VCONVOLVE VADD or VSUBTRACT operation It may also be used to threshold one part of an image differently from another The effect of a VTHRESHOLD operation is visible in VDISPLAY mode 2 It does not affect the associated grayscale image 260 AdeptVision Reference Guide Rev A Program Instruction Example cam virt 1 PARAMETER VIHRESHOLD cam virt 54 Threshold the current image at graylevel 54 VTHRESHOLD cam virt Related Keyword VEDGE program instruction AdeptVision Reference Guide Rev A VTHRESHOLD 261 V THRESHOLD System Parameter Syntax V THRESHOLD camera Function Set the camera grayscale value that separates black pixels from white pixels Usage Considerations Changing this parameter immediately affects the video output of the camera interface board assuming the VISION switch is enabled This is an array of parameters one for each virtual camera See the AdeptVision User s Guide for details on setting parameters Details This parameter sets the software threshold that is used to determine whether a camera pixel is to be interpreted as being white or bla
163. en returning results to the application program in millimeter coordinates for example the locations of edge transitions found by a ruler Perspective millimeters to simple millimeters This transformation corrects only for perspective distortion The user can use this to pre cisely place tools in cases where the vision system does not apply the transformation pmm to pix such as VWINDOW and graphics instructions This is the inverse of pmm to mm This transformation can be used to correct for perspective distortion in vision results when the vision system does not transform the coordinates with the transformation pix to pmm such as with blob or prototype locations Figure C 1 shows the three spaces or coordinate systems that have been discussed above and how the various transformations can be used to convert from one coordinate system to another pmm to pix PAM Mmmto pmm MM PIX Perspective Simple verae Vision corrected milimeter lt gt space millimeter pmm to mm space amp y scale space mm mm mm pix to pmm Figure C 1 Vision Coordinate Systems AdeptVision Reference Guide Rev A Appendix C Calibration Arrays The vision system uses pmm to pix to place rulers both linear and circular and the centers of arc finders All the other tools including line and point finders are placed using only the scale transformation
164. ensity The V EDGE TYPE parameter determines the edge operator used cross gradient or Sobel The V EDGE STRENGTH parameter is the edge threshold If the edge magnitude in the neighborhood of a pixel is greater than V EDGE STRENGTH the pixel is considered white Otherwise the pixel is considered black The resulting image consists of regions of intensity change V EDGE STRENGTH is also used as an edge sensitivity threshold for the inspection operators fine edge ruler type 5 VWINDOWL VFIND LINE VFIND POINT and VFIND ARC For example consider fine edge rulers As the grayscale pixels along the ruler are scanned the gradients in the neighborhood of each pixel are computed Then if the maximum gradient in the neighborhood is greater than V EDGE STRENGTH a search is started for the center of the edge along the ruler The position of the edge center computed with subpixel accuracy via interpolation is then reported by VRULER as the edge point As the value of V EDGE STRENGTH is lowered more edge points will appear in the image and will be reported by VRULER On the other hand increasing V EDGE STRENGTH will reduce the number of edge points To determine the best value of V EDGE STRENGTH disable the V BINARY system switch and enter display mode 2 so that the edges are visible in the Vision display window Then take a series of pictures each with VPICTURE cam 2 with a different value of VEDGE STRENGTH for each picture AdeptVis
165. enter Bottom most based on box center Top most based on box center Left most based on nearest box edge Right most based on nearest box edge V DAN 0 0 FPF WO Ne QO Bottom most based on nearest box edge o Top most based on nearest box edge 162 AdeptVision Reference Guide Rev A Program Instruction VLOCATE If the order parameter is omitted the value 0 is assumed which means no particular order is desired order values 3 through 6 select criteria based on the center of the object s bounding box For example an order value of 3 is a request for the object whose bounding box center is nearest the left side of the image Values 6 through 10 select criteria based on the object s closest edge For example an order value of 8 is a request for the object whose right edge is nearest the right side of the image When a nonzero value is specified for order in a VLOCATE instruction the vision system searches through the queue of processed objects to find the one that satisfies the requested order If the vision system has not completed processing of all the objects in the image the unprocessed objects are not considered in the order search If you want all the objects to be included execute a VWAIT program instruction before the VLOCATE to ensure that image processing has completed The real valued function VFEATURE 1 must be evaluated to determine if an object was located Memory is reserved fo
166. eptVision User s Guide for syntax details Details This parameter is useful for depalletizing applications where the camera sees only a portion of the pallet at one time In that case the four sides of the image will cut across that is clip some of the objects in the pallet into two portions a visible portion and an invisible portion The four sides of the image are defined by the parameters V FIRST COL V LAST COL V FIRST LINE and V LAST LINE or by an AOI definition If an object has been cut by the edge of the image and more than half of the object must be visible for it to be recognized the vision system may spend a lot of time proposing prototype matches to that object Only a portion of an object needs to be visible in order to propose a match but most or all of the object depending on the prototype s verify percentage and edge weights needs to be visible to verify the match V BORDER DIST defines an inner border within the outer border defined by the parameters V FIRST COL V LAST COL V FIRST LINE and V LAST LINE or by AOI definition Only the reduced image area inside the inner border is used during the match proposal step of image processing The inner border affects match proposals as described below If an object is completely outside the inner border and completely inside the outer border a match proposal is made because the object is completely visible That is the inner border has no effect
167. ere is a specific time during vertical blank when charges are transferred out of the pixels There can be no strobing during this time Therefore there is a no strobe region defined that prevents strobing If the picture is requested triggered during the no strobe time the strobe is postponed until just after the no strobe region Since this region is 1 25ms long there can be a worst case latency from picture request to image acquire of 1 25ms when using this mode Note also that if the V program containing the VPICTURE instruction uses an external signal to initiate the VPICTURE request a delay occurs between the external event and firing of the strobe See the description of the V IO WAIT parameter for information about this activation delay If V SYNC STROBE is set to 1 the strobe light is fired synchronously always at a fixed time with respect to the camera read out cycle Normally this time is during the read out of the 4th horizontal line Sometimes with CCD cameras firing a bright strobe light can cause a ghost image to appear in one field of the image This is caused by excessive light that leaks into the vertical shift registers This ghosting can be minimized or even eliminated by turning down the aperture of the camera lens to reduce the amount of light entering the camera If this is not possible the DEVICE instruction can be used to set the strobe output at line 20 By firing the strobe at this time ghosting never
168. erent especially with small angular ranges This is clarified below VFIND ARC types 2 1 and 0 are binary arc finders That is arcs are fit to binary edge points which are black to white or white to black transitions in the image Type 2 arc finders use the grayscale frame store This is the data visible in VDISPLAY mode 1 The values of the threshold parameters V THRESHOLD and V 2ND THRESH at the time the finder is executed are used to determine which pixels are black and white Therefore a different binary threshold can be used for each finder Type 1 arc finders use the raw binary frame store This is the data visible in VDISPLAY mode 2 This is binary or edge data depending on the setting of the V BINARY system switch Type 0 arc finders use the processed binary data produced by VWINDOW or VPICTURE in modes 1 or 0 This data is shown in VDISPLAY mode 3 Note that system parameters such as V MIN AREA and V FIRST COL are used during picture processing whereas they are not taken into account with type 1 or 2 finders Again the image data is binary or edge data depending on the setting of the V BINARY system switch Type 2 arc finders are the highest precision arc finders The edge points used in the least squares type of fit algorithm are found in the grayscale frame store with subpixel accuracy The image data used is visible in the Vision display window in VDISPLAY mode 1 The portion of t
169. erformed Otherwise the gap information is not retained and VGAPS results in the error message Information not available Example Before this program is called the parameter V LAST VER DIST must be nonzero a VPICTURE must be performed and a VLOCATE of the prototype must succeed The name of the prototype must be passed to the proto parameter of this program AdeptVision Reference Guide Rev A 135 VGAPS Program Instruction 136 The second argument to VLOCATE vloc must have been specified and then passed to this program This is a transformation variable that is assigned the location of the object PROGRAM label_gaps proto vloc ABSTRACT This program displays in the Monitor display window i information about all the gaps in a recognized prototype and puts a G at the center of each gap in the Vision display window r INPUT PARAMS proto name of the prototype to examine x vloc transformation variable that is assigned the location of the object LOCAL cang cnt d360_circum dx dy einfo enum LOCAL gaps gcx gcy glist gnum mid xta LOCAL type Get a list of all the prototype edges that have gaps VGAPS glist 0 TYPE Total verify percentage glist 1 Display a description of each prototype edge that has a gap Then display the gap ranges and label the gaps in the vision display window FOR cnt 1 TO glist 0 enum glist cnt
170. ering iteration should remove all the noise However if there are many noise edges near the true edge multiple iterations will be performed and some good edge points may be filtered out An example situation where many iterations would be performed is when the line finder tries to fit a line to a nonlinear edge such as an arc The allowable range for V MAX SD is 0 to 5 inclusive V MAX SD is set to 0 when V and AdeptVision systems are loaded into memory from disk Given a normal distribution of edge points 5 standard deviations should encompass all of the edge points so it will usually have the same effect as a V MAX SD value of 0 that is causing no edge points to be filtered out If edge filtering is desired VMAX SD should be in the range 1 to 2 for the best results in most situations Related Keywords VFIND ARC program instruction VFIND LINE program instruction V MAX PIXEL VAR system parameter AdeptVision Reference Guide Rev A 171 V MAX TIME System Parameter Syntax V MAX TIME camera Function Set the maximum time allowed for the vision system to analyze a region during object finding prototype recognition or OCR A value of 0 means that there is no time limit Usage Considerations A change to this parameter takes effect when a new region is analyzed This is an array of parameters one for each virtual camera See the AdeptVision User s Guide for details on setting parameters Details This pa
171. es are required is determined by the shape parameter See Details below angl ang2 Orientation or angular range of the area of interest The interpretation of these values as well as which values are required is determined by the shape parameter See Details below Details This instruction defines the shape size and orientation for most vision tools Figure 2 2 below shows the shape parameters that define rectangular or linear shaped tools Note that an area of interest can be defined that includes all elements but still can be used for shapes that do not require all elements For example the instruction VDEF AOI 2000 1 25 45 20 10 45 can be used by both point finders and linear rulers In the case of a point finder the tool will be centered at x 25 y 45 with a width of 20mm a height of 10mm and a rotation of 45 If the area of interest is used for a linear ruler the width specification will be ignored resulting in a ruler centered at x 25 y 45 witha 50 Adept Vision Reference Guide Rev A Program Instruction VDEF AOI length of 20mm and a rotation of 45 If you want to define an area of interest just for a ruler you have to leave a place marker for the ignored dimensions For example the following definition could be used for linear tools but not for tools requiring a full rectangle VDEF AOI 2000 1 25 45 20 0 45 A shape 3 definition accepts positive or negative values for dim3
172. es of a prototype are numbered starting at 1 VSHOW can be used to display a prototype with its edges numbered These are the edge numbers used with VEDGE INFO to reference the edges Each subprototype has its own edge numbering starting with 1 Again VSHOW can be used to display a subprototype with its edges numbered To use VEDGE INFO to get information about a region the V EDGE INFO switch must be enabled before the VPICTURE or VWINDOW instruction is executed Then the region must be VLOCATEd A second VPICTURE or VWWINDOW operation deletes all edge information from the previous VPICTURE or VWINDOW operation even if a different virtual camera number is specified When a different virtual camera is specified objects from the previous VPICTURE may still be VLOCATEd but no edge information is available After a region has been VLOCATEd VFEATURE item 27 provides the number of edges bounding the region not counting holes To get information about holes follow the same procedure for getting VFEATURE information about holes If V DISJOINT is true VEDGE INFO does not return edge information for regions that helped verify the recognition of a prototype Example 88 PROGRAM display_edges proto ABSTRACT This program displays in the Monitor display window 7 information about the edges of a prototype or subprototype The name of the prototype or subprototype is passed to the E program when it is called INPUT PAR
173. es whether or not to overlay the graphics memory on the image to be displayed Overlay values are 0 for no overlay 1 for overlay and 2 for static overlay See Details below xpan Optional real value variable or expression in the range 0 to 1023 specifying a pan in the x direction The default is 0 ypan Optional real value variable or expression in the range 0 to 511 specifying a pan in the y direction The default is 0 zoom Optional real value variable or expression indicating the magnification of the displayed image Acceptable values are 0 5 1 2 and 4 The default is 1 Adept Vision Reference Guide Rev A 79 VDISPLAY Details 80 Monitor Command and Program Instruction The zoom argument is encoded with a resolution as follows zoom zoom factor resolution 1000 Resolution Description 0 No change to the resolution Full resolution Half resolution Quarter resolution Ae WO N e Focus mode Modes are set and remembered individually for each virtual camera or for all virtual cameras if camera is 0 In all modes VDISPLAY has immediate effect if the vision system is not actively processing an image Otherwise the VDISPLAY mode will go into effect the next time a VPICTURE command or instruction is executed The following describes the vision display modes Mode 1 Grayscale camera output The live camera image is displayed in the Vision display window This is useful for positioning the ca
174. et to 1 for cross gradient when the V and AdeptVision systems are loaded into memory from disk The cross gradient operator is faster but the Sobel operator is a bit more robust The V EDGE STRENGTH system parameter is the threshold for both operators The following details are only for the curious they are not required for use Both operators are applied to every 3x3 neighborhood in the image window defined by the system parameters V FIRST COL V LAST COL V FIRST LINE and V LAST LINE Let the 9 grayscale pixels in a 3x3 neighborhood be labeled a through i a b c d e f g h i Then the cross gradient computed for the neighborhood is simply Edge MAX ABS a i ABS c g The Sobel computed for the neighborhood is more complex The vertical and horizontal edges are computed separately squared added together and the square root of the sum is computed The system performs this operation more efficiently than the following definition suggests 94 AdeptVision Reference Guide Rev A System Parameter V EDGE TYPE Dx a 2 d g c 2 f i Dy a 2 b c g 2 h i Edge SQRT Dx Dx Dy Dy 4 Dx and Dy are the edge magnitudes in the horizontal and vertical directions respectively Dx and Dy are more clearly depicted as convolutions with positive and negative weights 0 E 2 Dx 2 0 2 Dy 0 0 0 0 2 Fo
175. etermine edge numbers for the edge_num parameter in the VSUBPROTO instruction VSUBPROTO provides the requested information only if it refers to the most recent picture taken with a VPICTURE or VWWINDOW operation and the most recent object located with VLOCATE Also the V LAST VER DIST system parameter must have been nonzero when the VPICTURE was performed Otherwise the needed information is lost and VSUBPROTO produces the error message Information not available Examples The following examples show all the valid combinations of the input parameters for the instruction that is the parameters on the right of the equal sign For each example shown one or the other but not both of the output variables ver or vloc could be omitted Adept Vision Reference Guide Rev A 251 VSUBPROTO Program Instruction Consider all of the subprototype FLANGE ASIDE VSUBPROTO ver vloc FLANGE ASIDE Consider only edge 2 of the subprototype FLANGE BSIDE VSUBPROTO ver vloc FLANGE BSIDE 2 Consider only edge 16 of the latest prototype VSUBPROTO ver vloc 16 Related Keywords VDEESUBPROTO program instruction VGAPS program instruction VSHOW monitor command VSHOW program instruction V LAST VER DIST system parameter 252 AdeptVision Reference Guide Rev A Syntax VSUBTRACT Function Program Instruction VSUBTRACT cam type dmode dest_ibr srcl_
176. eters wide The rotation angle is 90 degrees Adept Vision Reference Guide Rev A 127 VFIND POINT Program Instruction dim2 67 33 dim3 40 dim4 25 angl 90 rect_ibr 4012 area of interest 4 virt frame buffer 12 VDEF AOI rect_ibr shape diml dim2 dim3 dim4 angl VFIND POINT cam pos data rect_ibr IF data 0 THEN TYPE Point found at F0 3 data 2 data 3 ELSE TYPE No point found END Related Keywords 128 VDEFAOI program instruction VFIND ARC program instruction VFIND LINE program instruction VRULERI program instruction V SHOW EDGES system switch Adept Vision Reference Guide Rev A System Parameter V FIRST COL Syntax V FIRST COL camera Function Set the number of the first column of pixels to be processed Usage Considerations A change to this parameter takes effect when the next VPICTURE or VWWINDOW is executed This is an array of parameters one for each virtual camera See the general description of parameters in the AdeptVision User s Guide for syntax details Details This parameter together with V LAST COL is used to set the range of camera pixel columns that are processed during VPICTURE and VWWINDOW operations Columns outside the range specified are ignored by the vision system VDEF AOI is the preferred method for defining a processing border for VPICTURE and VWINDOW This parameter must be assigned an integer value in
177. etion including all recognition attempts before processing of the next region is begun When scenes consist of multiple objects and V DISJOINT is enabled the vision system uses more memory and may take more time to process the image The V DISJOINT and V TOUCHING switches affect the interpretation of the how_many parameter to the VPICTURE and VWINDOW instructions That parameter specifies the maximum number of objects the vision system should try to recognize V DISJOINT affects how the how_many parameter applies to the image as a whole whereas V TOUCHING affects how it applies to each region in the image If V DISJOINT is enabled the number of objects specified by how_many is the most that will be recognized in the entire image If V DISJOINT is disabled there is no limit on how many objects will be recognized in the entire image Adept Vision Reference Guide Rev A 77 V DISJOINT System Switch The following table summarizes the relationship between V DISJOINT V TOUCHING and the how_many parameter to VPICTURE and VWINDOW Number of Objects Number of V TOUCHING V DISJOINT per Region Objects per Scene Off Off 1 No limit Off On 1 how_many On Off how_many No limit On On how_many how_many Example The two images in Figure 2 5 illustrate an object a fork that appears to the vision system to be split into two parts Thus the V DISJOINT switch must be enabled to recognize the object
178. ettings of bits 3 get hole mode and 6 get subproto mode The order of the models in the system is determined by the order in which they are created using VTRAIN or loaded into the system from disk Get proto hole 0 versus Get subproto 1 mask value 32 NOTE Get proto hole and Get subproto are not applicable for ObjectFinder models This bit is ignored if bit 2 get specific mode or 3 get hole mode is set Get proto hole 0 Requests information about a prototype or hole See bit 2 Get next 0 Adept Vision Reference Guide Rev A 227 VSHOW Program Instruction Get subproto 1 This mode requests information about a subprototype In order for this mode to be used a prototype first has to be VSHOWed The subsequent get subproto VSHOWS refer to that prototype Each get subproto VSHOW returns one subprototype of the prototype In this mode the proto_name parameter must be a string variable VSHOW in get hole or get subproto mode refers to the prototype most recently VSHOWed regardless of which program task executed the VSHOW instruction Consequently there is possible confusion when more than one program task executes VSHOW instructions Thus for predictable operation with these request modes applications should be organized to have only one program task execute VSHOW instructions If the edge_num parameter is nonzero one or more edge numbers are shown These edge numbers may be needed fo
179. evel histogram which is shown in blue Peaks in the red curve are the good thresholds the higher the peak the better the threshold 28 AdeptVision Reference Guide Rev A Monitor Command and Program Instruction VAUTOTHR The highest peak in the average contrast curve is not always the best threshold however For example specular reflections on shiny parts may produce a high peak in the curve The associated threshold would separate only the specular reflections from the rest of the image instead of separating the object from the background Since these specular reflections are usually the brightest part of the image their intensity will be located at the high end of the histogram To automate the filtering out of high average contrasts near the extreme ends of the histogram VAUTOTHR uses the system parameter V MIN AREA to trim the ends That is if the number of pixels in the histogram above or below a peak is less than V MIN AREA the peak is discarded This is done because there would be regions in the image with areas less than V MIN AREA if one of these discarded peaks were used as a threshold An effective strategy for auto thresholding is first to choose manually the best threshold and then later have VAUTOTHR search for a threshold within a limited range of the current threshold To choose manually the best threshold execute VAUTOTHR as a monitor command with no parameters The entire intensity range is considered and the resu
180. eywords that are similar or are frequently used in conjunction with this instruction are listed here AdeptVision Reference Guide Rev A 19 V 2ND MOMENTS System Switch Syntax V 2ND MOMENTS camera Function Enable computation of the best fit ellipse for each region in the image Usage Considerations The system switches V CENTROID and V BOUNDARIES must also be enabled in order for the second moments of inertia to be computed A change to this switch takes effect when the next VPICTURE command or instruction or VWINDOW instruction is executed This is an array of switches one for each virtual camera See the general description of switches in AdeptVision User s Guide for syntax details Details V 2ND MOMENTS enables the computation of the second moments of inertia for regions Based on the second moments and centroid of each region a best fit ellipse is fit to the region The dimensions of the best fit ellipse are available from the VFEATURE function after a VLOCATE instruction has succeeded VFEATURE provides the ellipse major and minor radii and the angle of the ellipse major axis The center of the ellipse is at the centroid of the region which is also available from VFEATURE The areas and positions of holes are not considered during the computation of moments The VSUBTRACT HOLE system switch does not affect moments However if you want these statistics for holes you should enable V HOLES disable V DISJOI
181. f tolerance To identify all the gaps in a prototype or subprototype the prototype must first be VLOCATEd Then use VGAPS with an edge_num value of 0 This returns into data a list of the edges that have gaps in them Then use VGAPS again with these edge numbers to find where the gaps are within each particular edge A gap is defined by a pair of distances in millimeters from the start of the edge line or arc The portion of the edge between the pair of distances is a nonverified gap As their numbering indicates the edges of each region in a prototype are ordered in a clockwise direction The distances in the data array returned by VGAPS use the same direction If a pair of distances in data indicating a gap are 0 and 5 2 then the gap starts at the first corner of the edge moving clockwise around the region extending into the edge 5 2 millimeters Circular edges begin at 0 degrees the point on the circle that is farthest right in the Vision display window If only some of the edges of the prototype or subprototype are of interest use the VSHOW monitor command to display all of the edge numbers for the prototype Then use VGAPS with each edge number of interest VGAPS provides the requested information only if it refers to the most recent picture taken via VPICTURE and the most recent object located via VLOCATE Also the V LAST VER DIST system parameter must be nonzero when the VPICTURE operation is p
182. ffect Example Disable object recognition for all virtual cameras do only boundary analysis ENABLE V BOUNDARIES DISABLE V RECOGNITION Related Keywords VLOCATE program instruction V SHOW RECOG system switch Adept Vision Reference Guide Rev A 211 VRENAME Monitor Command Syntax VRENAME new_name old_name Function Rename a prototype or subprototype Usage Considerations This command may be used even while a main control program is executing Parameters new_name The new name of a prototype or subprototype If a subprototype is being named this parameter must have the form namel name2 entered without quotation marks where namel is the name of the prototype containing the subprototype name2 Either namel name2 or both may be new old_name The name of an existing prototype or subprototype See above for the syntax of subprototype names Examples Rename the prototype VRENAME SPADE SHOVEL Rename the subprototype VRENAME GASKET NOTCH GASKET INDENT Details The type of model being supported is determined by the name which has been trained as either a prototype or an ObjectFinder model The name is checked first against the lists of prototypes for backward compatibility Therefore you should not use the same name for a prototype model and an ObjectFinder model Related Keywords VDELETE monitor command and program instruction VSHOW monitor com
183. file_spec Smodel_name Store in a disk file selected or all vision prototypes and their subprototypes Optical Character Recognition OCR fonts or correlation templates Smodel_name V STROBE camera Enable the firing of a strobe light in synchronization with taking pictures VPICTUREs VSUBPROTO ver_percent subproto_name edge_num Determine the percentage of an edge or subprototype that was verified during recognition Also this instruction can have the prototype position refined based on only a subprototype or a single edge producing an adjusted location for the prototype trans_var 298 VSUBTRACT cam type dmode dest_ibr srcl_ibr src2_ibr Subtract two binary or grayscale images V SUBTRACT HOLE camera Determine whether or not hole areas are to be subtracted from region areas V SYNC STROBE camera Select synchronous or asynchronous firing of a strobe light when a picture is taken that is when a VPICTURE is executed VTHRESHOLD cam src_ibr Threshold a grayscale image producing a binary image type dmode dest_ibr V THRESHOLD camera Set the camera grayscale value that separates black pixels from white pixels V TOUCHING camera Determine whether or not objects may be touching in the image VTRAIN Sprototype height ang VTRAIN Sprototype ibr Initiate training of the prototype whose name is specified shape cx cy width VTRAIN value VTRAIN cam mode arg
184. fined convolution and a value in the range 17 to 32 performs a user defined convolution see the description of VDEF CONVOLVE Optional real valued expression specifying the display mode for this operator The choices are 1 no draw 0 erase 1 draw solid 2 complement 3 draw dashed and 4 complement dashed The default is 1 draw solid Optional integer value specifying the image buffer region that will receive the result of the convolution Image buffer regions specify both a virtual frame buffer and a size see the description of VDEF AOD Optional integer value specifying the image buffer region to convolve The image buffer regions AOI must have been defined with a VDEF AOI instruction NOTE If only one image buffer region is specified both image buffer regions default to the specified image buffer region If neither image buffer region is specified both default to the currently selected full frame AdeptVision Reference Guide Rev A 41 VCONVOLVE Program Instruction The AOIs for convolutions greater than 3x3 are limited to 504 pixels in width If an AOI wider than 504 is specified the image buffer region is reduced based on the center of the image buffer region Details A convolution is a neighborhood pixel operator that may be used to smooth an image or extract edges The effect of a convolution is visible in VDISPLAY mode 1 As a first experiment with the VCONVOLVE instruction type VDISP 1 and
185. for good thresholds These values must be in the range 0 to 127 inclusive The value of start must be less than the value of end By default start is 0 and end is 127 Optional array into which the threshold values are placed The first value in the array is the number of good thresholds found n The next n values in the array are the thresholds in order of goodness see below for details Optional integer value that identifies the first array element to be defined in array Zero is assumed if the index is omitted If a multiple dimension array is specified only the right most index is incremented as the values are assigned Optional real value variable or expression interpreted as an integer that specifies the image buffer region to threshold The image buffer region specifies both an AOI and a frame store See the definition of VDEF AOL VAUTOTHR computes a histogram for the image currently in memory and automatically determines good threshold values With automatic threshold determination the vision system can adapt to gradual changes in lighting over time The sample program given below for example could be called once every 30 minutes to adjust the threshold VAUTOTHR finds optimum thresholds by computing the average contrast in the image produced by every threshold The average contrast for every possible threshold 0 to 127 is shown as a red curve overlaying the gray l
186. for three different values of V MAX PIXEL VAR PARAMETER V MAX PIXEL VAR 0 75 PARAMETER V MAX PIXEL VAR 1 5 PARAMETER V MAX PIXEL VAR 3 0 Figure 2 9 Effects of VIMAX PIXEL VAR Parameter 168 Adept Vision Reference Guide Rev A System Parameter V MAX PIXEL VAR Related Keywords VFIND ARC program instruction VFIND LINE program instruction V MAX VER DIST system parameter VPICTURE monitor command and program instruction VTRAIN monitor command VTRAIN program instruction VWINDOW program instruction AdeptVision Reference Guide Rev A 169 V MAX SD System Parameter Syntax V MAX SD camera Function Set the distance in units of standard deviation from the fit line or arc beyond which edge points should be filtered out Usage Considerations A change to this parameter takes effect when a VFIND LINE or VFIND ARC program instruction is performed This is an array of parameters one for each virtual camera See the AdeptVision User s Guide for details on setting parameters Details 170 This parameter is used only during execution of the line and arc finders VFIND LINE and VFIND ARC If V MAX SD is nonzero the finder filters out edge points that are more than V MAX SD standard deviations from the fit line or arc edge and then the finder refits the edge This filtering is disabled if V MAX SD is zero In more detail the filtering operation is as follows First the finder either VFIND
187. formed and is automatically drawn into when instructions such as VRULER and VWWINDOW are executed See the description of mode 3 above particularly that regarding the erasure time An overlay value of 2 is like VDISPLAY mode 4 That is the vision display window is static modified only with user graphics instructions When acquiring images in single field mode a zoom value of 4 is not allowed The image is already horizontally zoomed by 2x so it will display correctly The x and y pan values allow you to scan the entire frame store regardless of which areas have valid image data A single frame grabbed image uses 640 x 480 pixels leaving the remaining 1024 x 512 frame buffer untouched However operations such as VCOPY and VCONVOLVE may write to any portion of the frame store Correlation templates are also stored in unused portions of the frame store Examples 82 Turn on special display mode for virtual camera 4 VDISPLAY 4 3 Enable live video input with static overlay VDISPLAY 1 2 Turn on live thresholded mode for all virtual cameras VDISPLAY AdeptVision Reference Guide Rev A Monitor Command and Program Instruction VDISPLAY Related Keywords VDEEFAOI program instruction VPICTURE monitor command and program instruction VSELECT program instruction V SHOW BOUNDS system switch V SHOW EDGES system switch V SHOW GRIP system switch V SHOW RECOG system switch V
188. fter the first stage of processing the computation of run lengths is complete Therefore the run lengths are computed on the binary edge image which is the result of VEDGE see Appendix B for details on how vision run lengths are generated In each case above the choice of edge operation to be performed cross gradient or Sobel is determined by the value of the system parameter V EDGE TYPE And the edge strength threshold is given by the VEDGE STRENGTH system parameter Examples Enable binary image processing for virtual camera 1 but not for virtual camera 2 ENABLE V BINARY 1 DISABLE V BINARY 2 Adept Vision Reference Guide Rev A 33 V BINARY System Switch Related Keywords 34 VAUTOTHR monitor command and program instruction VPICTURE monitor command and program instruction VEDGE program instruction V EDGE STRENGTH system parameter V EDGE TYPE system parameter V THRESHOLD system parameter V 2ND THRESH system parameter AdeptVision Reference Guide Rev A System Parameter V BORDER DIST Syntax V BORDER DIST camera Function Define an image border reduction in pixels to mask out regions clipped by the image border Usage Considerations This parameter takes effect when the next VPICTURE command or instruction or VWINDOW instruction is executed This is an array of parameters one for each virtual camera See the general description of parameters in the Ad
189. ful when the VPICTURE is waiting for a fast digital input interrupt signal before starting to acquire an image see below 4 If wait 1 VPICTURE will wait until the first stage of processing run length encoding is complete The above exceptions do not apply if planning for prototype recognition is performed In that case VPICTURE always waits until the frame grab is done before completing If the V IO WAIT system parameter is set to 1 a VPICTURE waits for a fast digital input interrupt signal before starting to acquire an image Consequently except when the VPICTURE wait parameter is 1 the program containing the VPICTURE instruction will wait an unlimited amount of time until the input signal is received See the description of the V IO WAIT parameter for details The future frame grab provides the ability to pregrab an image and hold it for processing later mode 1 This is useful when the vision system is still processing a previous image and the next one is in view of the camera The vision system can process only one image at a time Instead of making a robot or manufacturing process wait for the current image processing to complete mode 1 may be used to pregrab the next image and thus allow the robot or process to proceed With a future frame grab no operation may be performed on the frame grabbed image except for a VPICTURE in mode 0 However a future frame grab can be converted into a quick frame gra
190. fy the pixel tolerance to use when determining boundary coincidence VLOAD file spec Load vision models ObjectFinder models prototypes Optical Character Recognition fonts or correlation templates from a disk file VLOAD lun Sfile_spec Load vision models ObjectFinder models prototypes Optical Character Recognition fonts or correlation templates from a disk file VLOCATE camera trans_var Identify and locate an object in the scene mode order name AdeptVision Reference Guide Rev A Appendix A AdeptVision Quick Reference V MAX AREA camera Set the maximum area above which the vision system ignores regions V MAX PIXEL VAR camera During a VFIND LINE or VFIND ARC operation this parameter specifies the maximum pixel distance from the fit edge beyond which edge points may be filtered out During boundary analysis this parameter sets the maximum pixel deviation allowed when fitting lines and arcs to region edges V MAX SD camera Set the distance in units of standard deviation from the fit line or arc beyond which edge points should be filtered out V MAX TIME camera Set the maximum time allowed for the vision system to analyze a region during object finding prototype recognition or OCR A value of 0 means that there is no time limit V MAX VER DIST camera Set the pixel tolerance for determining boundary coincidence during the verification of prototype to image matches V MIN AREA came
191. gaps gnum 2 gt gaps gnum 2 1 S Line Arc TYPE Cl Corner einfo 4 einfo 5 END Related Keywords VDEESUBPROTO program instruction VSHOW monitor command VSHOW program instruction VSUBPROTO program instruction V LAST VER DIST system parameter Adept Vision Reference Guide Rev A 137 VGET AOI Program Instruction Syntax VGET AOI array i aoi Function Return the definition of an area of interest Usage Considerations The area of interest must have been defined with a VDEF AOI instruction Parameters array Variable name identifying the array to receive the AOI definition i Optional integer value specifying the starting array index aoi Real valued expression identifying the area of interest Details If the VGET AOI instruction specifies a valid AOI number the following information is returned in array arrayli Returns the shape number if definition is valid returns 1 if definition is not valid array i 1 Dimension 1 of AOI array i 2 Dimension 2 of AOI array i 3 Dimension 3 of AOI array i 4 Dimension 4 of AOI array i 5 Angle 1 of AOI array i 6 Angle 2 of AOI Example Change dim3 of aoi 3 to 37 4 VGET AOI def 3000 IF def 0 lt gt 1 THEN VDEF AOI 3000 def 0 def 1 def 2 37 4 def 4 def 5 END Related Keyword VDEF AOI program instruction 138 AdeptVision Reference Guide Rev A Synta
192. gly For example a camera lens with a shorter focal length may be needed to increase the field of view Or increase the distance between the camera and objects to be viewed Also the vision tools used to locate the object in the image will need a wider search space along the direction of travel of the object Example Select asynchronous strobing for camera 1 V SYNC STROBE 1 0 Related Keywords VPICTURE monitor command and program instruction V IO WAIT system parameter V STROBE system switch 16 4cm per 640 pixels 0 01 cm pixel 0 1 mm pixel 2 0 0167 second 1000 mm second 0 1 mm pixel 167 pixels AdeptVision Reference Guide Rev A 259 VTHRESHOLD Program Instruction Syntax VTHRESHOLD cam type dmode dest_ibr src_ibr Function Threshold a grayscale image producing a binary image Parameters cam Optional integer specifying the virtual camera that will supply values for V THRESHOLD and V 2ND THRESHOLD type Optional integer identifying the image type 1 Same result as type 1 see below Allows the switch V BACKLIGHT to be passed as the type value O Reverse the output polarity substitute a binary bit value of 0 for each 1 and a binary bit value of 1 for each 0 1 Default No change to output polarity dmode Optional real valued expression specifying the display mode for this operator The choices are 1 no draw 0 erase 1 draw solid 2 complement 3 draw dashed and 4 compl
193. h options should be disabled The mask values for each option are Mask Pull Down Item Value New example 1 Verify percent 2 Effort level 4 Min max areas 8 Limit position 16 Edge weights 32 Assign cameras 64 Details VTRAIN is used to create a new object prototype or to modify an existing prototype If the prototype exists the user may show the vision system a new instance or change the prototype s virtual camera associations effort level edge weights verification threshold minimum maximum areas or position constraints A window may be specified when creating a new object prototype or training another instance of a prototype The window can be used to limit the portion of the image that is considered The window specification is the same as that for the VWINDOW instruction which is used to perform object recognition within windows Once VTRAIN has been initiated the mouse is used to interact with the system Training initiated with a VTRAIN instruction can be terminated by aborting the program that contains the instruction See the AdeptVision User s Guide for a description of the training process Exa mple Initiate vision training for a prototype named flange VTRAIN flange Adept Vision Reference Guide Rev A 269 VTRAIN Program Instruction Related Keywords VDEF AOI program instruction VLOCATE program instruction VSTORE program instruction 270 Adept Vision Reference Guide Rev A Program
194. h the AdeptVision Enhanced VXL Interface option 4x4 or smaller takes 9ms larger kernels take 17ms Also the width of the AOI is limited to 508 pixels for kernels larger than 4x4 Convolutions numbered 1 to 16 are predefined and may not be redefined by the user The predefined convolutions are as follows SS O O O O O O O O Q 1 3x3 flat average 0 0 14 14 14 0 0 3 3x3 vertical edge 0 0 14 14 14 0 0 0 O O O O O 0 0 0 14 14 14 0 0 0 0 25 75 25 0 0 o oo Oo Oo Oo Oo gt O YO O O o oo Oo O Oo Oo O O O OG O O OS gt O O O o O O O O QO O O O gt O O O O O O O O Q 2 3x3 gaussian average 0 0 5 15 4 3x3 horizontal edge 0 0 AdeptVision Reference Guide Rev A 0 0 15 47 15 0 0 0 0 75 0 0 0 5 15 0 0 25 0 ooo O Oo Oo Oo 0 ooo ooo ooo Oo Oo Oo Oo gt O O O O O O 57 VDEFECONVOLVE 58 SS D O OG O Q O QS O O SO O o lt r Sy O O Q O O a a a a a O 5 3x3 diagonal edge 7 5x5 flat average 0 O a 01 a a Aa 9 7x7 vertical edge 1 0 0 O a 01 a a A 0 G O O O O O 0 O a a a a A 1 5 13 17 13 5 1 Program Instruction SO O O O Q O a a a a a O 11 14 11 G O O O oa o oo Oo Oo O Oo SS O i O O Q O ao O OG DD O O 1 1 oS O 0O OG QO D ON U Po O CO O a gt 1 6
195. he weighted boundary of the prototype that was verified If following a VSHOW instruction this is the verify percentage required for recognition of an instance of the prototype ITEM 10 Following a VLOCATE this is the area of the region in camera pixels If the VSUBTRACT HOLE system switch has been enabled the region area has the areas of holes subtracted from it Otherwise they are included In either case the total area of the holes is available as VFEATURE item 40 Note that this is the area of the region not the area of the prototype instance Following a VSHOW this feature is the area of the prototype ITEMS 11 and 12 After VLOCATE only Region ID is a count of the number of regions encountered Instance ID is a count of the number of objects recognized and regions not recognized These ID numbers start at 1 and count up The distinction between region and instance is for touching and overlapping parts If two objects touch and both are recognized they account for 1 region ID increment and 2 instance ID increments If a region consists of multiple objects and the objects are not recognized the vision system has no way of knowing how many objects there are so the region ID and instance ID counts are both incremented by 1 If the vision V DISJOINT system switch has been enabled there may be more regions than instances because a single prototype may span two or more regions ITEMS 13 to 16 Min_X Max_X Min_Y and Max_Y defi
196. he first element to be defined in data 1 and locs respectively Zero is assumed for any index that is omitted If a multiple dimension array is specified only the right most index is incremented as the values are assigned 182 Adept Vision Reference Guide Rev A Program Instruction VOCR font_num Real valued expression that specifies the font to be recognized or verified The value must be in the range 1 to 99 Sexpected Optional string containing the text expected to be found within the OCR window This string is ignored if op is 3 This string is required for all other values of op ibr Integer value specifying the image buffer region within which to search for characters for recognition or verification Image buffer regions specify both a size and a frame store see the description of VDEF AOD shape Optional real valued expression indicating the shape of the window Currently the only choice is 1 for a rectangular window cx cy Real valued expressions specifying the center coordinates of the window in millimeters dx Real valued expression specifying the width of the window in millimeters dy Real valued expression specifying the height of the window in millimeters ang Optional real valued expression specifying the orientation of the window in degrees The default is 0 degrees Details This instruction performs the OCR functions of text verification or recognition When text is verified
197. he image through which VFIND ARC searches for edge points is shaped like a ring or a ring segment This area is defined by the area of interest definition or the instruction parameters for the estimated arc center xc yc estimated arc radius r the search range rr and the angular range ang0 angn AdeptVision Reference Guide Rev A Program Instruction VFIND ARC The mode bits for the VFIND ARC mode parameter are defined as follows they all default to zero Bit 1 LSB Dark inside 0 versus light inside 1 mask value 1 If the inner part of the arc or circle is light this bit should be set Otherwise this bit should be clear Based on this bit VFIND ARC searches for edges with a particular sign pixel transitions that are light to dark or dark to light Bits 2 3 Center only 0 radius only 2 both 4 mask value 6 Center only 0 In this mode the given radius dim3 of the area of interest definition is assumed to be precise and only the arc s circle center is computed Radius only 0 In this mode the given center first and second elements of the area of interest definition is assumed to be precise and only the arc s radius is computed Both 4 In this mode both the arc s circle center and radius are fit The given arc center dim1 and dim2 elements of the area of interest definition and radius dim3 of the area of interest definition are used only to define the a
198. he vision processor must be idle and vision model training must not be active for this instruction to be executed The V LAST VER DIST system parameter must have been nonzero when the last VPICTURE or VWINDOW was performed to make the necessary information available to the VSUBPROTO instruction The VSUBPROTO instruction refers to the object most recently VLOCATEd regardless of which program task executed the VLOCATE instruction Consequently for predictable operation applications should be organized to have only one program task execute VLOCATE instructions Parameters 250 ver_percent Optional real valued variable that is to be assigned the percentage of the subprototype or edge that was verified 0 to 100 trans_var Optional transformation variable to be assigned the refined location of the prototype object NOTE ver_percent and or trans_var must be specified subproto_name Optional string expression specifying the name of the subprototype in the form namel name2 where namel is the name of the prototype last VLOCATEd and name2 is the name of the subprototype edge_num Optional real valued expression that specifies the edge number of the prototype or subprototype for which information is requested If subproto_name is specified edge_num is optional and refers to an edge of the subprototype using its edge numbering If edge_num is omitted or has the value 0 the entire subp
199. hes at every pixel location results in a much faster correlation tool However certain templates such as the ones with low contrast or with skinny features do not permit skipping VTRAIN MODEL analyzes each template during training to determine a default skip pattern and a default depth of hierarchy The determination criteria places higher priority on robustness than speed resulting in a more conservative default depth and skip pattern Being conservative means a slower tool but with less variation in execution time and less chance of not finding the object The search time limit for levels 0 4 is controlled by the V MAX TIME parameter A value of 0 means that there is no time limit for the search The default values can be overridden by specifying the desired depth and skip pattern using input parameters max_depth and mode respectively However not using the recommended depth and skip pattern may result in mismatches or increased variability of execution times If the depth parameter is greater than 0 a hierarchical search is performed The image and template are first subsampled by averaging and mode number of reduced images are produced A search is then made of the most reduced image and the location with the highest correlation match is selected Next the mode 1 reduced image is searched in close proximity to the location with the best match This process is iterated until the final target region image is s
200. his slows down the execution of VOCR somewhat Planning for OCR needs to be performed one time after a font has been trained or loaded from disk using VLOAD If the font being used has not been planned the VOCR instruction automatically initiates planning Also the VTRAIN MODEL instruction may be used to force the vision system to plan for OCR VOCR returns a variety of information to satisfy different inspection requirements The information returned by the four modes of operation selected with the op parameter is explained below Each operation mode returns match scores in the range 0 to 100 A score of 0 means no correlation A score of 100 means a perfect match Scores of 100 are uncommon To earn this a character would have to be the exact average of all its trained examples Operation 0 Quick Verification In this mode of operation the bounded areas in the image buffer region are compared against the given expected text expected on a 1 to 1 basis This is the fastest mode of operation Only three values are returned in the array data data i 0 Number of character regions found and analyzed data i 1 Average score of Sexpected characters verified data i 2 Minimum score of Sexpected characters verified Operation 1 Robust Verification Short Output This is a more robust form of verification than Quick Verification op 0 but it is slower Even though VOCR knows what text to expect it co
201. i The definition of similar radii is liberal depending on the standard deviations of the radii of the prototype arcs and the effort levels of the prototypes The V IBOUNDARIES system switch overrides the V FIT ARCS switch That is if V BOUNDARIES is disabled arcs will not be fit regardless of the setting of V FIT ARCS Similarly if the V MAX PIXEL VAR system parameter is 0 both line fitting and arc fitting are automatically disabled Related Keywords V BOUNDARIES system switch VFEATURE real valued function VPICTURE monitor command and program instruction V RECOGNITION system switch Adept Vision Reference Guide Rev A 131 V GAIN System Parameter Syntax V GAIN camera Function Set the gain for the incoming video camera signal Usage Considerations Changing this parameter immediately affects the video output of the camera interface board This is an array of parameters one for each virtual camera See the general description of parameters in the AdeptVision User s Guide for syntax details Details The V GAIN parameter works with the V OFFSET parameter to select the incoming analog video gain and offset respectively V GAIN multiplies scales the video signal whereas V OFFSET shifts translates the video signal Before adjusting the values of VOFFSET and V GAIN you should take a picture compute the histogram and study the video data displayed in the Vision display window To take the pic
202. ibr src2_ibr Subtract two binary or grayscale images Parameters cam type dmode dest_ibr srcl_ibr src2_ibr Details Selects the virtual camera to supply threshold values used in creating the binary image Optional integer value indicating the type of subtraction 1 for binary 2 for grayscale averaging or 3 for grayscale subtraction The default is 1 binary Optional real valued expression specifying the display mode for this operator The choices are 1 no draw 0 erase 1 draw solid 2 complement 3 draw dashed and 4 complement dashed The default is 1 draw solid Optional integer value specifying the image buffer region that will receive the result of the image subtraction Image buffer regions specify both a size and a frame store see the description of VDEF AOI If specified the resulting image is stored in this image buffer region Otherwise the result replaces the source image buffer region srcl_ibr Integer values specifying the image buffer regions to subtract The image buffer regions AOIs must have been defined with a VDEF AOI instruction sre2_ibr will be subtracted from srcl_ibr Image subtraction may be useful to subtract fixtures from an image or to inspect a part by comparing it with the stored image of a golden part Of course precisely repeatable placement of the parts is necessary in this latter case Image subtraction is also useful in combination with other image t
203. ide Rev A 151 V LAST COL System Parameter Syntax V LAST COL camera Function Set the number of the last column of pixels to be processed Usage Considerations A change to this parameter takes effect when the next VPICTURE or VWWINDOW is executed This is an array of parameters one for each virtual camera See the general description of vision parameters in the AdeptVision User s Guide for syntax details Details This parameter together with V FIRST COL is used to set the range of pixel columns that are processed during VPICTURE and VWINDOW operations Columns outside the range specified are ignored by the vision system VDEF AOI is the preferred method for defining a processing border for VPICTURE and VWINDOW The value of the parameter V LAST COL must be greater than the value of V FIRST COL and less than or equal to 640 The parameter V LAST COL is set to its maximum value when the V and AdeptVision systems are loaded into memory from disk Example Process all image data up to the 256th column of pixels PARAMETER V LAST COL 256 Related Keywords VDEF AOI program instruction V FIRST COL system parameter V FIRST LINE system parameter V LAST LINE system parameter 1 This is the maximum setting The effective value of V LAST COL is limited by the virtual frame allocation made with the DEVICE instruction or the CONFIG_C utility 152 AdeptVision Reference Guide Rev A System Parameter V LAST LINE
204. ientation in the image Multiple windows in the same image may be processed Windows may overlap or even be nested inside each other Many small windows in an image may be processed in less time than it would take to process the entire image The virtual camera parameter cam selects the group of prototypes system switches and system parameters to use when processing the window This also selects the calibration array to be accessed The VWINDOW instruction is essentially a repicture VPICTURE 0 inside a window Almost all of the system switches and parameters apply Thus recognition may be performed in one window connectivity statistics computed in another etc However some system features are applied only during the actual acquisition of an image that is when a VPICTURE instruction acquires an image The following system switches and parameters are not referenced by VWINDOW V GAIN V OFFSET V BINARY V THRESHOLD and V 2ND THRESH The V BORDER DIST parameter is considered only with orthogonal windows It is ignored for rotated windows Adept Vision Reference Guide Rev A 285 VWINDOW Program Instruction A VPICTURE operation must be initiated before window processing so that an image is available for analysis Any VPICTURE mode may be used except for future frame grab mode 1 Quick frame grab mode 2 is the recommended mode Either of the two frame stores may be used see the VSELECT program instruction
205. ification see step 7 below that is required when more than one prototype feature is associated with a feature class Even if there is only one feature associated with a class confirmation will be used if the prototype is complex having many edges so that mismatches will be detected before the expensive verification step The need for confirmation is determined during training when feature classes are chosen Associated with each prototype feature is a list of the confirming features that if present in the image at the same relative position would distinguish it from other features Nearby features are chosen as confirming features when appropriate because they are more likely to be visible not occluded along with the feature to be confirmed Verify Match Given a single prototype to image match proposal this step verifies the presence of the prototype in the image The prototype boundary representation is transformed to the image via the 2 D translation and rotation proposed in step 6 and a search is made for image edges that align with the prototype edges The close enough test is to within a user specified tolerance V MAX VER DIST Although the signs of the edges black to white versus white to black must correspond the edge types do not So lines verify arcs and vice versa If enough of the prototype boundary is present in the image the match is accepted Enough is defined by a weighted threshold based on the edge weights and
206. ilation respectively Types 3 through 7 are undefined reserved for future built in operations Types 8 through 16 are user definable Type 9 is actually predefined to be Life a simulation of cellular organisms popularized years ago in Scientific American but it can be redefined by the user Binary morphological operations are applied to every 3x3 pixel neighborhood in the binary image Based on the binary pixel values in a neighborhood and the operation to be performed the center binary pixel value may be changed See the VDEF MORPH instruction in this manual for details on how the operation is performed The effect of a VMORPH operation is visible in VDISPLAY mode 2 It does not affect the associated grayscale image The erosion operation clears binary pixels that have at least one neighboring pixel with the value 0 The dilation operation does the opposite setting all pixels that have at least one neighboring pixel with the value 1 The smaller the area of the image to be processed the faster VMORPH executes Examples VDEF AOI 2000 1 100 100 50 50 VMORPH 2 2011 Dilate the binary image twice VMORPH 2 2011 VMORPH 1 2011 Then erode the image twice VMORPH 1 2011 Related Keywords VDEFAOI program instruction VDEEMORPH program instruction AdeptVision Reference Guide Rev A 181 VOCR Program Instruction Syntax VOCR cam op dmode data i
207. imes to retrieve the entire boundary description Additional memory and some extra processing time is spent storing the edge information so if the information is not needed this switch should be disabled This switch is disabled when the V and AdeptVision systems are loaded into memory from disk 90 Adept Vision Reference Guide Rev A Example System Switch Save edge info for virtual camera 1 ENABLE V EI DGE INFO 1 Related Keyword VEDGE INFO program instruction AdeptVision Reference Guide Rev A V EDGE INFO 91 V EDGE STRENGTH System Parameter Syntax V EDGE STRENGTH camera Function Set the edge threshold for grayscale image processing and fine edge rulers Usage Considerations A change to this parameter takes effect when the next VPICTURE VEDGE VRULER VWWINDOWI VFIND LINE VFIND POINT or VFIND ARC instruction is executed This is an array of parameters one for each virtual camera See the general description of parameters in the AdeptVision User s Guide for syntax details Details 92 This parameter is used by various vision functions the first of which is to threshold the edges during grayscale image processing Grayscale image processing occurs in response to a VEDGE instruction or a VPICTURE request when the V BINARY system switch is disabled With grayscale edge processing edges are extracted from the image based on local gradients local changes in image int
208. in the Vision display window Overlays the models on the found instances AdeptVision Reference Guide Rev A 115 VFIND LINE Program Instruction Syntax VFIND LINE VFIND LINE Function cam pos dmode effort type data i ibr cam pos dmode effort type data i 1 xc yc length width angle Fit a straight line to an image edge within a window Usage Considerations If the calibration being used includes correction for perspective distortion the correction is applied to the returned values describing the fit line Adept recommends that you use the first syntax Parameters cam pos dmode effort type data 116 Optional real valued expression that specifies a virtual camera number The default is 1 The camera number is used to determine which V EDGE STRENGTH parameter to use Optional real valued expression indicating the starting point in the search window 1 dark side 0 middle guideline 1 light side The default value of pos is 0 Optional real valued expression specifying the display mode for this operator The choices are 1 no draw 0 erase 1 draw solid 2 complement 3 draw dashed and 4 complement dashed The default is 1 draw solid Optional real valued expression indicating the effort level to use when fitting the line The effort level can range from 1 least effort to 100 maximum effort The default is 50 Optional real valued expre
209. inder Tool Start Position and Polarity AdeptVision Reference Guide Rev A 119 VFIND LINE Program Instruction Figure 2 8 shows a finder tool and the instructions that created the tool shape 1 diml 20 4 dim2 25 6 dim3 15 dim4 12 angl 0 rect_ibr 3000 22 jaoi 3 frame buffer 22 VDEF AOI rect_ibr shape diml dim2 dim3 dim4 angl VFIND LINE 1 1 1 data rect_ibr dim3 gt Figure 2 8 Sample Line Finder Tool The area searched for edges always has its sides aligned along the horizontal vertical or diagonal directions Consequently the VFIND LINE window is a parallelogram not necessarily a rectangle The guideline can have any orientation but the sides of the search window at the ends of the guideline are always horizontal vertical or diagonal In the two search windows shown below imagine that the guideline in the left hand window is 20 degrees off vertical and the guideline in the right hand window is 30 degrees off vertical The windows have different shapes because the window automatically changes orientation from vertical to diagonal when the guideline is at 22 5 degrees since that is halfway between 0 and 45 degrees This is the worst case shape distortion since it is near the transition point 120 Adept Vision Reference Guide Rev A Program Instruction VFIND LINE The effort level specifies how dense the scan is along the guideline Only a few edge points are s
210. index is incremented as the values are referenced 64 AdeptVision Reference Guide Rev A Program Instruction VDEFGRIP w Array of real values defining the widths of the rectangles in millimeters num_fngrs number of values must have been stored in sequential array elements h Array of real values defining the heights of the rectangles in millimeters num_fngrs number of values must have been stored in sequential array elements Details One to four grip models may be defined per prototype Each grip model consists of one to five rectangles in any orientation representing the gripper fingerprints around the prototype For simple two finger grippers two rectangles are adequate Odd shaped grippers may be approximated by defining multiple overlapping rectangles Grip models are automatically stored and restored with the prototype model when the VSTORE and VLOAD commands are issued The grips defined for a prototype may be seen in the Vision display window via the VSHOW command The clear grip test is automatically performed whenever a prototype is recognized and grippers are defined for it The results of the test are available via the VFEATURE function and are shown in the Vision display window with a graphics display mode when the V ISHOW GRIPS switch is enabled Grips are tested in the order of their numbering 1 through 4 Once a grip is found to be clear further testing is halted because only one clear grip is requ
211. ined elements of the input arrays are automatically assumed to be zero Subsequent VGETCAL instructions will retrieve arrays that are defined completely NOTE The user is discouraged from setting any of the array values unless explicitly requested to do so by Adept documentation or personnel The descriptions here are intentionally terse Table 2 5 Elements of VGETCAL VPUTCAL Scaler Calibration Array Index Name Value or Meaning 0 calibrated Calibration status 0 gt Uninitialized 1 gt Initialized for calibration 2 gt Okay calibrated 1 cam num Physical camera number 2 method Calibration method and link mounting used The value of this element is determined and used by Adept software See the documentation for the Adept calibration utilities for specific values For camera only calibrations method 1 link num Link number of the robot that the camera is mounted on 4 Maintained for backward compatibility should not be used in 11 0 or later systems 5 6 7 not used Must be 0 8 not used Must be 0 or 1 9 12 not used 13 flags Bit field Not set Set 1 HPSnotused HPS used 2 Full frame Field acquire 3 Normal Synchronous shuttered camera 4 Normal Asynchronous shuttered camera Remaining bits reserved must be 0 14 xy ratio Camera pixel height width converts X to Y Read only returned by VGETCAL but not used by VPUTCAL 15 x scale Horizontal scale factor mm pixel
212. ing conditions For arc rulers the angles are accurate to about half a pixel 216 Adept Vision Reference Guide Rev A Program Instruction VRULERI For type 2 or 3 rulers you can include the mags parameter to request that the gradient magnitudes for each of the edges found be returned into that array Positive gradients indicate a dark to light edge and negative gradients indicate a light to dark edge The absolute values of these gradient magnitudes are compared against the V EDGE STRENGTH parameter to determine if pixels are edge candidates For ruler types 2 1 0 2 and 3 that is edge rulers the starting element in the array datali contains the number of transitions found Therefore the number of elements in the array defined after a VRULERI is data i 2 The maximum number of transitions that VRULERI will locate and report is 100 or the value of the maxcnt parameter Any additional transitions are ignored If a ruler is requested that lies entirely off the image zero transitions are returned If a ruler lies only partially on the image transitions are detected only for the section of the ruler that is in the image When an arc ruler is clipped the maxcnt parameter is ignored and all the transitions along the ruler are returned The coordinates of the end points of a linear edge ruler or the center of an arc ruler may be outside of the image but they must be within a 1000 pixel r
213. ing the coordinate of the lower left corner of the image to be read The coordinate is in pixel units and must be within the image The default is 1 1 the lower left corner on the Vision display window dx dy Optional real valued expressions specifying the width and height of the image to read in pixel units The image dimensions are listed in the table above If the width or height of the image window being read exceeds the image dimensions on the top or right the dimensions are automatically reduced and the new dimensions are stored in the header string If not specified ax and ay default to the full image size For grayscale VGETPICs dx is limited to 512 pixels Details 142 This instruction reads a rectangular window of an image and stores it into a two dimension string array Unless the data is packed see below each string in the array holds 128 pixels As many strings as needed are filled for each row four is the maximum required There is one row of strings for each row of the image window being read A header string is stored in pic r c normally the 0 0 location which contains pertinent information about the stored image See below Type 1 VGETPIC reads the image in the frame store currently selected See the VSELECT program instruction This includes both the grayscale data lower 7 bits of each byte and the binary or edge data high bit of each byte A type 2 VGETPIC reads only the
214. instruction This is an array of switches one for each virtual camera See the AdeptVision User s Guide for details on setting switches There are two strobe outputs Strobe 1 is used for physical cameras 1 and 3 Strobe 2 is used for physical cameras 2 and 4 See the Adept MV Controller User s Guide for cable details Details 248 When V STROBE is enabled for a virtual camera the strobe light is fired when a VPICTURE operation is performed for that camera Strobe lights are usually used to take pictures of moving objects without getting a blurred image For example if objects are riding on a moving conveyor belt a trigger mechanism such as a photoelectric cell may be connected to a digital input line that signals a V application program when a VPICTURE instruction should be executed NOTE For the fastest most consistent response to an external event the fast digital input interrupt line should be used For more information see the description in this manual of the V IO WAIT system parameter The timing of the strobe signal within the picture acquisition and with respect to the VPICTURE request depends on the system parameter V SYNC STROBE This parameter determines if the strobe is fired synchronously or asynchronously with respect to the camera read out cycle and if there is to be a reset of the camera s vertical drive signals See the description in this manual of V SYNC STROBE for details The following constraints
215. ion Enhanced VXL Interface option Parameters cam Optional integer value specifying a virtual camera The V THRESHOLD and V 2ND THRESH values associated with the virtual camera are used when the scale parameter is 0 5 see Details scale Optional real valued expression specifies magnification of the copy The acceptable values are 0 5 1 and 2 The default is 1 This feature requires the AdeptVision Enhanced VXL Interface option dmode Optional real valued expression specifying the display mode for this operator The choices are 1 no draw 0 erase 1 draw solid 2 complement 3 draw dashed and 4 complement dashed The default is 1 draw solid lut Optional integer in the range 1 to 133 maximum the actual maximum value is set with the DEVICE instruction that specifies a defined look up table The table is defined with a VDEF LUT instruction and specifies a replacement value for each possible graylevel value See the description of VDEF LUT This feature requires the AdeptVision Enhanced VXL Interface option dest_ibr Integer value specifying the image buffer region to receive the copied image Image buffer regions specify both a size and a frame store see the description of VDEF AOD seo br Integer values specifying the image buffer region to copy The image buffer regions s AOI must have been defined with a VDEF AOI instruction AdeptVision Reference Guide Rev A 43 VCOPY Program Instruction NO
216. ion Reference Guide Rev A System Parameter V EDGE STRENGTH This parameter must be assigned an integer value in the range 0 to 127 inclusive The parameter is set to 20 when the V and AdeptVision systems are loaded into memory from disk Example Use 15 as the edge threshold for all cameras PARAMETER V EDGE STRENGTH 15 Related Keywords V BINARY system switch VEDGE program instruction VFIND ARC program instruction VFIND LINE program instruction VFIND POINT program instruction VPICTURE program instruction VRULERI program instruction VWINDOWI program instruction V BINARY system parameter V EDGE TYPE system parameter AdeptVision Reference Guide Rev A 93 V EDGE TYPE System Parameter Syntax V EDGE TYPE camera Function Determine the type of edge operator to use cross gradient or Sobel when a VPICTURE instruction is performed Usage Considerations A change to this parameter takes effect when the next VPICTURE instruction is executed This is an array of parameters one for each virtual camera See the general description of parameters in this manual for syntax details Details This parameter determines the method of edge extraction for VPICTURE when the V BINARY system switch is disabled If the V IEDGE TYPE parameter value is 1 a cross gradient operator is applied to the grayscale image Otherwise if the parameter value is 2 the Sobel operator is used The parameter is s
217. ion can be used to extract information about areas of an image Adept Vision Reference Guide Rev A 217 VRULERI Program Instruction For run length type 0 linear rulers the more horizontal the ruler the faster it is processed Exactly horizontal rulers are a special case and process fastest their speed is independent of their length Horizontal or vertical type 2 and 1 rulers are slightly faster than those with random angle orientations With the exception of exactly horizontal type 0 rulers type 4 1 rulers are typically the fastest rulers For fine edge types 2 and 3 rulers orientation has no effect but they are about half as fast as a run length binary ruler The speed of graylevel type 1 rulers is dependent only on their length and shape When no region oriented information is needed and single pixel precision is sufficient the fastest cycle time can be achieved by using a quick frame grab VPICTURE in mode 2 followed by type 1 rulers These raw binary rulers are very fast Enable the V BINARY system switch for maximum speed of a VPICTURE in mode 2 If object recognition is not needed but the filtering effects of parameters such as V MIN AREA are needed the time delay before rulers are processed can be minimized by disabling the VBOUNDARIES system switch The following comments on perspective distortion may be of interest to the expert user If the calibration being used includes correction for per
218. ion system is not in an idle state Example Take a quick picture VPICTURE 2 Compute and display the histogram VHISTOGRAM Related Keywords VDEFAOI program instruction VAUTOTHR monitor command and program instruction VWINDOWI program instruction V GAIN system parameter V OFFSET system parameter AdeptVision Reference Guide Rev A 147 V HOLES System Switch Syntax V HOLES camera Function Enable or disable the accounting of interior features in all objects Usage Considerations A change to this switch takes effect when the next VPICTURE command or instruction or VWINDOW instruction is executed The setting of this switch is ignored by the vision system when the V DISJOINT switch is enabled This is an array of switches one for each virtual camera See the general description of switches in the AdeptVision User s Guide for syntax details Details If the V HOLES switch is enabled the system keeps track of the holes in all objects whether or not the objects have been recognized The description of a hole is available by executing the VLOCATE instruction followed by use of the VFEATURE function Holes are described by their area bounding box and their parent child and sibling relationships with other holes Also if an object has been recognized the VFEATURE function reports which holes helped verify the presence of the object in the image Determining hole locations may be
219. ired The clear grip algorithm checks each of the rectangles defining the grip until one is found to be not clear or until all are found to be clear A rectangle is clear if it covers only background colored pixels where the background is defined by the system V BACKLIGHT switch As a consequence a rectangle completely enclosed in a hole that is the color of the background is considered clear Regions smaller than V MIN AREA that merge into the background assume the color of the background Thus these noise regions do not cause the clear grip test to fail Exa mple In the following example gripper fingerprints are defined for the prototype named CASTING Two rectangles are used to model the gripper both 30 millimeters tall and 7 5 millimeters wide The rectangles are located plus and minus 25 millimeters from the prototype center along the X axis and 10 millimeters up from the center along the Y axis The rectangles are oriented orthogonally with respect to the X and Y axes AdeptVision Reference Guide Rev A 65 VDEFGRIP Program Instruction SET rects 0 TRANS 25 10 0 SET rects 1 TRANS 25 10 0 wids 0 7 5 wids 1 7 5 hts 0 30 hts 1 30 VDEFGRIP CASTING 1 TRUE 2 rects wids hts Related Keywords VFEATURE real valued function VSHOW monitor command VSHOW program instruction V SHOW GRIP system switch 66 AdeptVision Reference Guide Rev A Program Instru
220. is an array of parameters one for each virtual camera See the AdeptVision User s Guide for details on setting parameters Details Regions that contain fewer than V MIN AREA number of pixels are ignored by the vision system during object recognition clear grip testing and standard ruler measurements This parameter should be set large enough so that the vision system does not waste time analyzing noise spots and small enough so that no region of interest is discarded This parameter must be assigned an integer value in the range 1 to 1 048 576 1024 1024 inclusive However the value must be less than or equal to the value of VMAX AREA and it must be greater than or equal to the value of the parameter V MIN HOLE AREA The parameter V MIN AREA is set to 16 when the V and AdeptVision systems are loaded into memory from disk Example Ignore regions that are smaller than 25 pixels in area PARAMETER V MIN AREA 25 Related Keywords V MAX AREA system parameter V MIN HOLE AREA system parameter 1 The effective maximum of VIMAX AREA is the area of the virtual frame buffer Virtual frame buffer sizes are specified with the DEVICE instruction Adept Vision Reference Guide Rev A 175 V MIN HOLE AREA System Parameter Syntax V MIN HOLE AREA camera Function Set the minimum area below which the vision system ignores holes Usage Considerations A change to this parameter takes effect when a new region is analyze
221. is larger than V MAX VER DIST V LAST VER DIST reaches out to features of the object that may be out of tolerance The program instructions VSUBPROTO and VGAPS provide final verification information on selected individual edges or subprototypes of the prototype AdeptVision Reference Guide Rev A 173 V MAX VER DIST System Parameter This parameter must be assigned a real value in the range 1 to 16 inclusive V MAX VER DIST should always be greater than VMAX PIXEL VAR The parameter V MAX VER DIST is set to 3 when the V and AdeptVision systems are loaded into memory from disk Example PARAMETER V MAX VER DIST 3 5 Related Keywords 174 VGAPS program instruction V LAST VER DIST system parameter V MAX PIXEL VAR system parameter V RECOGNITION system parameter VSUBPROTO program instruction Adept Vision Reference Guide Rev A System Parameter V MIN AREA Syntax V MIN AREA camera Function Set the minimum area below which the vision system ignores regions Usage Considerations Normally changes to this parameter take effect immediately filtering out any small regions that close off However when clear grips have been defined changes to this parameter do not go into effect until a new VPICTURE is executed The effects of changes to this parameter are visible in the Vision display window only in a special display mode Regions smaller than V MIN AREA will still appear in display modes 1 and 2 This
222. ith VGETPIC VQUEUE Display object information for any objects queued in the vision system awaiting retrieval by VLOCATE instructions VQUEUE Return the number of objects in the vision system queue V RECOGNITION Enable or disable prototype recognition by the vision system VRENAME Rename a prototype or subprototype VRULERI Obtain edge information or graylevels along a line or circular arc in the current image VSELECT Select one of the two grayscale frame stores for processing or display and optionally select a virtual camera and its calibration to be associated with the frame store VSHOW List the defined prototypes or display a vision prototype a subprototype or a specific prototype edge in the Vision display window Edge numbers are optionally shown 12 Adept Vision Reference Guide Rev A Chapter 1 AdeptVision Keyword Summary Table 1 2 AdeptVision Keyword Summary Continued Keyword Description VSHOW Display a vision prototype or subprototype and make information about it available through the VFEATURE real valued function V SHOW BOUNDS Enable the special display of the lines and arcs fit to the boundaries of regions V SHOW EDGES Enable the special display of edges both the primitive edges that are fit to the boundaries of regions and the edge points that are found by the finders VFIND LINE VFIND ARC and VFIND POINT V SHOW
223. izes the combinations Number of Objects Number of V TOUCHING V DISJOINT per Region Objects per Scene Off Off 1 No limit Off On 1 how_many On Off how_many No limit On On how_many how_many NOTE If clear grip tests have been defined for the prototypes see VDEFGRIP objects located that do not have a clear grip are not counted toward the how_many limit Thus more than how_many objects may be found and placed in the vision queue see VQUEUE Examples VPICTURE Find at most 3 objects with virtual camera 1 VPICTURE Find as many objects as possible with camera 4 VPICTURE 1 Grab a frame for later processing VPICTURE 0 1 Find as many objects as possible in the image being held from virtual camera 7 Initiate picture processing with a monitor command producing the following output VPICTURE Running Objects to be recognized CASTING FLANGE CRANK BRACKE T KKK 10 15 AN PS 25 30 ll Only 5 examples trained Only 1 example ignored Only 3 examples trained 1 The V TOUCHING switch is automatically considered to be enabled whenever the V OVERLAPPING switch is enabled AdeptVision Reference Guide Rev A 199 VPICTURE Monitor Command and Program Instruction The following code shows how to perform inspections at frame rates 30 frames per second For simplicity this program performs a single window inspection on e
224. k and sets the binary bit to black l t3 3 FOR x 0 TO 50 Zero out indices where the binary lut x 0 bit is 0 and the graylevel value END is lt 50 FOR x 128 TO 238 Zero out indices where the binary lut x 0 bit is 1 and the graylevel value END is lt 50 AdeptVision Reference Guide Rev A 67 VDEELUT FOR x lut x EN D FOR x EN VDEF LUT lut3 lut lut x D Related Keywords 68 Program Instruction Leave remaining values unchanged DL to 127 Xx 239 to 255 Xx VCOPY program instruction DEVICE program instruction Adept Vision Reference Guide Rev A Program Instruction VDEEMORPH Syntax VDEF MORPH mode type array dx dy Function Define a binary morphological operation Usage Considerations VDEF MORPH is intended only for experienced users It is not critical to the operation of the vision system The VISION switch must be enabled the vision processor must be idle and vision model training must not be active for this instruction to be executed Parameters mode Not currently used type Integer value in the range 9 to 16 inclusive This is the number of the user definable morphological operation that is being defined array Array of length 512 that defines the morphological operation Each element in the array must have a zero or nonzero value dx dy Not currently used Details Binary morphological operations are nonlinear transf
225. lement 3 draw dashed and 4 complement dashed The default is 1 draw solid Integer value specifying the image buffer region that will receive the result of adding the source image buffer regions Image buffer regions specify both a virtual frame buffer and an area of interest see the description of VDEF AOD Integer values specifying the image buffer regions to add The AOIs src2_ibr must have been defined with a VDEF AOI instruction The virtual frame stores specified for these two image buffer regions must be in different physical frame stores The VADD operation adds two images The images considered are identical areas of interest within two frame stores The areas are defined using VDEF AOI instructions If the areas of interest are not the same size the larger areas of interest are shrunk to the same dimensions as the smallest but each keeps its original center location Each pixel in the area of interest for one image is added to the corresponding pixel in the other image and the result is stored in the corresponding pixel of the destination image buffer region The smaller the area of interest to be processed the faster the addition executes AdeptVision Reference Guide Rev A 25 VADD Program Instruction VADD type 1 is binary addition The two binary images are simply OR d together pixel by pixel In comparison when the VSUBTRACT instruction is used in binary mode the images are exclusive OR d The underlying g
226. les 246 The following instruction stores all the prototypes from the vision system into a file named OBJECTS VS on the default system disk Logical unit number 5 is associated with the operation That logical unit number could be used to check for any error during the operation see below VSTORE 5 objects The following example stores a single font in the file FONT VS The number of the font is defined by the real variable font num which is used to construct the name of the font in the form FONT_n VSTORE 5 font vs FONT_ ENCODE 10 font num AdeptVision Reference Guide Rev A Program Instruction VSTORE The following example stores the prototypes smd1 and quad in a file named PROTOS VS on the hard disk C Logical unit number 6 is associated with the operation and is used to check for a successful completion Spnamel VSTORE smdl 6 C PROTOS Spnamel quad IF IOSTAT 6 lt 0 THEN TYPE C1 VSTORE failure ERROR IOSTAT 6 C1 HALT END Related Keywords VLOAD monitor command VLOAD program instruction VSTORE monitor command Adept Vision Reference Guide Rev A 247 V STROBE System Switch Syntax V STROBE camera Function Enable the firing of a strobe light in synchronization with taking pictures VPICTUREs Usage Considerations A change to this switch takes effect with the next VPICTURE command or
227. lines and arcs are fit to within a user specified tolerance V MAX PIXEL VAR so image smoothing is controlled Secondly the Adept Vision Reference Guide Rev A 303 Appendix B Classify Features result is further data reduction especially when the parts are circular or have circular holes Studies of industrial parts have shown that the majority of parts are prismatic and rotational in shape so their 2 D images are naturally represented by lines and arcs The final justification for fitting lines and arcs as opposed to fitting only lines is that they provide a greater variety of edge and corner types making recognition easier Classify Features This step classifies all connected edge edge pairs where the edges are the lines and arcs from step 4 and associates the pairs with the feature classes of the known prototypes The feature classes are chosen during training based on the objects the user wants the vision system to recognize Generically there are only a few types of corner classes line line line arc arc line arc arc and circle In addition there are the wildcard variations line any any line arc any and any arc The any refers to an edge that may belong to a different object because two parts touch or overlap Although there are only a few generic types of feature classes each of the feature classes of a prototype have specific parameterized definitions An acceptable minimum and maximum included angle th
228. ling V SHOW FEATS will decrease processing time because arc and line fitting takes a significant amount of time Related KeywordS V SHOW BOUNDS system switch 232 AdeptVision Reference Guide Rev A System Switch V SHOW GRIP Syntax V SHOW GRIP camera Function Enable the special display of clear grip tests Usage Considerations A change to this switch takes effect when the next region is displayed in the Vision display window with a graphical display mode in effect This is an array of switches one for each virtual camera See the AdeptVision User s Guide for details on setting switches Details When V SHOW GRIP is enabled clear grip tests are shown in a special display mode for example VDISPLAY mode 3 That is each rectangle of a grip definition is displayed as it is tested Of course this is performed only when gripper positions have been defined for the prototype and the prototype has been recognized Grips are tested in the order of their numbering 1 through 4 as defined with VDEFGRIP instructions Once a grip is found to be clear further testing is halted because only one clear grip is required Each grip is modeled by one to five rectangles The algorithm for testing for clear grips checks each of the rectangles until one is found to be not clear or until all are found to be clear Consequently when a grip is not clear the operator may see only some of the rectangles that define the grip Example Show
229. ling relationships Following a VLOCATE in the normal find object mode or a VSHOW instruction in get prototype mode the hole number is always 0 and the parent number is always 1 The parent in this case is the background Following a VLOCATE in the special find hole mode or a VSHOW in get hole mode the hole number is an integer value 1 or greater All of the holes in the object are numbered including holes in holes The parent number of a hole is the parent s hole number Thus if an object has one hole and the hole has one hole in it the first hole s hole number is 1 and its parent number is 0 The innermost hole has a hole number of 2 with a parent number of 1 ITEM 23 After a VLOCATE this is the number of the virtual camera with which the object was located After a VSHOW this is the number of a virtual camera associated with the prototype VFEATURE items 30 and 31 are bit masks representing all cameras associated with the prototype AdeptVision Reference Guide Rev A Real Valued Function VFEATURE ITEM 24 After VSHOW only The effort level assigned to a prototype during training ITEM 25 After VSHOW only The color black or white that the prototype had during training ITEM 26 After VSHOW only The number of prototype samples taught during training This is a measure of the reliability of the prototype ITEM 27 The number of bounds in the prototype VSHOWed o
230. lts are displayed Then overlay the results on the live binary image by executing the monitor command VDISPLAY 0 1 Finally try each of the optimum thresholds shown by changing the V THRESHOLD parameter and decide on the best setting The frame store must contain a valid picture Otherwise an error results See the VSELECT program instruction Example The following code computes a new threshold for the given virtual camera The search range is t 20 to t 20 where t is the current threshold LOCAL t tmin tmax VPICTURE cam 2 Take a quick picture t PARAMETER V THRESHOLD cam The current threshold Search range is 20 around the current threshold tmin MAX 0 t 20 Make sure it s gt 0 tmax MIN 127 t 20 Make sure it s lt 127 VWAIT Wait for image processing VAUTOTHR 0 tmin tmax thrs Compute best threshold IF thrs 0 gt 1 THEN If got one use it PARAMETER V THRESHOLD cam thrs 1 END Adept Vision Reference Guide Rev A 29 VAUTOTHR Monitor Command and Program Instruction Related Keywords 30 VDEEAOI system parameter VHISTOGRAM monitor command and program instruction V 2ND THRESH system parameter V MIN AREA system parameter V THRESHOLD system parameter Adept Vision Reference Guide Rev A System Switch V BACKLIGHT Syntax V BACKLIGHT camera Function Define which color black or white is
231. lution that is being defined array Array of dimensions 7x7 defining a convolution Each element in the array must be in the range 128 to 127 Pj Optional integers specifying the starting array index values The default value of i and j is 0 Details A convolution is a neighborhood pixel operator that may be used to smooth an image or extract edges The effect of a convolution is visible in VDISPLAY mode 1 The VCONVOLVE program instruction performs convolutions The convolution is applied to each pixel in the grayscale frame store a The convolution is applied to a single pixel by centering the convolution on the pixel so that the 7x7 array of convolution terms overlays the 7x7 pixel neighborhood of the pixel b Each term of the convolution is multiplied by the pixel value that it overlays c The 49 products are summed and divided by 128 56 AdeptVision Reference Guide Rev A O om O O OQ O O GO O QO O Program Instruction VDEECONVOLVE d The absolute value of the result is stored in the second image memory for the center pixel The exception is in convolution 14 where it has a signed output with zero at graylevel 64 All convolutions are 7x7 in size However smaller convolution sizes such as 3x3 may be defined inside a 7x7 matrix as shown below The speed of VCONVOLVE depends on the system and the kernel On the standard system the time is roughly 9ms for each nonzero element in the kernel Wit
232. m switch V CENTROID system switch V DISJOINT system switch V HOLES system switch VLOCATE program instruction V MIN MAX RADII system switch V PERIMETER system switch AdeptVision Reference Guide Rev A V 2ND MOMENTS V 2ND MOMENTS Necessary switches Attach to the vision window and select graphics scaling in real world millimeters Indexes of VFEATURE function for centroid and best fit ellipse dimensions Special display Take a picture with virtual Camera 1 and no recognition Wait for image processing to complete for graphics instr Locate anything in the image If a region was found Get centroid Cx Cy Get min max radii of ellipse Angle of major axis Draw major axis cy dy Draw minor axis cy dy Locate next region in the image 21 V 2ND THRESH System Parameter Syntax V 2ND THRESH camera Function Set a second threshold for use during binary image processing Usage Considerations Changing this parameter immediately affects the video input to the binary frame buffers assuming the VISION switch is enabled This is an array of parameters one for each virtual camera See the general description of parameters in AdeptVision User s Guide for syntax details Details 22 This is an optional second threshold that may be used with the primary threshold V THRESHOLD to separate black pixels from white ones V
233. m the VPUTCAL operation 1 The load area subroutine for loading calibration data can be found in the file LOADAREA V2 on Adept Utility Disk 2 See the AdeptVision User s Guide for information on use of this subroutine 308 Adept Vision Reference Guide Rev A Appendix C Calibration Arrays Calibration Amays The instructions VGETCAL and VPUTCAL have three real valued array Wh parameters scalers pmm to pix and pix to pmm The one dimensional array scalers contains the X scale and Y scale information described earlier By itself this array is sufficient for the scale transformation The arrays pmm to pix and pix to pmml are optional 3 by 3 arrays These arrays contain two dimensional homogeneous transformations that contain millimeters pixel scaling correction for perspective distortion and a small translation and rotation The array pmm to pix transforms millimeter coordinates on a single plane in the workspace into pixel coordinates in the image The array pix to pmm does the reverse transformation These transformations have the form B C F H 1 Elements A B D and E encapsulate scaling and rotation Elements C and F are the translations in X and Y respectively Elements G and H correct for perspective distortion the smaller these numbers the less perspective
234. mand VSHOW program instruction VTRAIN monitor command and program instruction 212 AdeptVision Reference Guide Rev A Program Instruction VRULERI Syntax VRULERI cam type dmode maxcnt edge dir data i mags j ibr VRULERI cam type dmode maxcnt edge dir data i mags j 1 x0 y0 len ang VRULERI cam type dmode maxcnt edge dir data i mags j shape cx Cy radius ang0 angn Function Obtain edge information or graylevels along a line or circular arc in the current image Usage Considerations If the calibration being used includes correction for perspective distortion the correction is applied to both the placement and results of the ruler Adept recommends the first syntax Parameters cam Optional real valued expression indicating the virtual camera number to use for selecting the V EDGE STRENGTH parameter that identifies strong edges for a fine edge type 2 or fine pitch type 3 ruler The default camera is 1 type Optional real valued expression specifying the type of ruler 2 dynamic binary 1 raw binary 0 run length binary default 1 graylevel 2 fine edge 3 fine pitch dmode Optional real valued expression indicating the display mode to use when displaying the ruler 1 no draw 0 erase 1 draw solid 2 complement 3 draw dashed and 4 complement dashed The default is 1 draw solid maxcnt Optional real valued expression specifying the maximum number of values to re
235. me for a prototype model and an ObjectFinder model NOTE When no model names are specified VSTORE stores all the models in memory to the disk file There is no way to globally store only the ObjectFinder models If you want to do this you must store each model individually by specifying the model name It will assume all prototypes if no names are given Like the VSTORE monitor command this instruction saves the specified prototypes fonts or templates in the given disk file Note however that prototypes fonts and templates may not be mixed in a file Thus all the model names must be prototypes or they must all be fonts or they must all be templates If no model names are specified all the prototypes currently in system memory are stored in the file If the model name FONT_0 is specified all the defined fonts are stored in the file Similarly if the model name is TMPL_0 all the defined templates are stored When prototypes are stored any subprototype definitions are stored along with them As the file is created the prototype or font information is written in a special format for later recall with the VLOAD command or instruction The IOSTAT real valued function can be used after this instruction to determine if any error occurred during the disk operation NOTE The application program must not have attached the logical unit since the VSTORE instruction automatically attaches and detaches the logical unit Examp
236. mera focusing and setting the aperture Also note that the system parameter settings V GAIN and V OFFSET affect the live camera image displayed This is the initial display mode NOTE If no virtual camera number is specified with modes 1 and 0 the image from the camera most recently referenced is displayed Mode 0 Mode 1 Mode 2 Thresholded camera output The Vision display window shows the live thresholded output of the camera This mode is primarily useful for setting the threshold for binary image processing The effects of the parameters V THRESHOLD and V 2ND THRESH are both visible See the Note above Frame grabbed grayscale With this display mode one of the two grayscale frame stores is displayed After a VPICTURE operation the frame acquired is displayed Otherwise VSELECT may be used to select the frame to be displayed Frame grabbed edge or binary With this mode one of the two binary edge frame stores is displayed Depending on whether the Adept Vision Reference Guide Rev A Mode 3 Mode 4 Mode 5 Monitor Command and Program Instruction VDISPLAY V BINARY system switch was enabled or not when the VPICTURE was last executed this display mode shows a frozen binary or edge image The VSELECT instruction may be used to select the frame to be displayed Special display This is the graphics display mode All user graphics histograms VSHOW of prototypes training images and the special display g
237. meter must be a string variable which is filled in with the name of the object found Find particular 1 If this bit is set to one a particular object is sought The name of the object to be found must be given The name may be given if an unknown region is sought Default Find any mode 0 Find object 0 versus Find hole 1 mask value 4 NOTE Find hole does not apply to the ObjectFinder tool Find object 0 This is the normal mode of operation where recognized objects or unrecognized regions are located Find hole 1 In this mode the VLOCATE is a request for hole information about the object last located This mode provides a way to inspect for unexpected holes in objects In order for this find hole mode to be used an object first has to be VLOCATEd in the normal find object mode The subsequent find hole VLOCATESs refer to the object found Each find hole VLOCATE returns one hole of the object The information provided by the VFEATURE function is sufficient to reconstruct the parent child and sibling relationships between the holes in case there are holes within holes AdeptVision Reference Guide Rev A 161 VLOCATE Program Instruction If two or more objects are touching or overlapping they have the same set of holes because they are all part of the same outer region Consequently the same hole may be VLOCATEd in find hole mode multiple times once after each object is recognized The duplicate hole desc
238. meters V MAX SD and V MAX PIXEL VAR control the filtering process VFIND LINE filters edge points if VMAX SD is nonzero Otherwise filtering is disabled V MAX SD specifies the number of distance standard deviations from the fit line beyond which edge Adept Vision Reference Guide Rev A 121 VFIND LINE Program Instruction points should be discarded V MAX PIXEL VAR specifies the maximum pixel distance below which no edge points should be discarded See the description of the V MAX SD system parameter for more information on the iterative filtering process The edge points found by VFIND LINE are displayed in the Vision display window when the system switch V SHOW EDGES is enabled Edge points used in the final fit of the line are displayed in white Edge points discarded during the filtering process are displayed in gray Note that displaying the edge points is computationally expensive so it should be enabled only for investigative purposes Items data i 7 data i 8 and data i 9 are returned only when edge filtering is enabled that is when V MAX SD is nonzero Item data i 9 is the percent of edge points filtered out This percentage ranges from 0 up toward 100 It can never be 100 however because the filtering process always stops before filtering out too many points The execution time for VFIND LINE is proportional to the effort level multiplied by the length of the guideline Also execution time i
239. most significant of which is that confirmation is not planned for Therefore during recognition all proposals go on to verification This method can be costly since verification is a time consuming operation Adept Vision Reference Guide Rev A 201 VPLAN FINDER Monitor Command and Program Instruction Input para meters using virtual cameras The following settings are required V MIN LEN vc 20 Percentage of features by length to ignore For Type O normal planning V MIN LEN is ignored For Type 1 quick planning V MIN LEN vc is the number of features to use for planning For example a value of 20 means to plan with the 20 longest features The following settings are suggested V MAX VER DIST vc 5 0 202 AdeptVision Reference Guide Rev A Syntax VPUTCAL Function Program Instruction VPUTCAL camera scalers i pmm to pix j k pix to pmm 1 m to cam Load vision calibration data from 1 to 3 arrays Usage Considerations The VISION switch must be enabled the vision processor must be idle and vision model training must not be active for this instruction to be executed Parameters camera scalers pmm to pix pix to pmm 1 Jr to cam Details Real valued expression that specifies the virtual camera Real array of calibration values see below Optional two dimensional real array that defines the millimeter to pixel transformation matrix If this array is specified the array
240. mpares each bounded area in the image buffer region with all the characters in the font The top two character matches for each bounded area are then compared to the expected character The top two matches or picks are the two with the highest match scores VOCR then returns data indicating whether all of the expected characters were first picks or at least first or second picks If all of the expected characters were first picks and the font is complete the text is guaranteed to be readable Even if the expected characters were all first or second picks the text was probably readable Some characters in the font such as 1 and 1 may be very similar and thus open to misreading by VOCR whereas a human would use AdeptVision Reference Guide Rev A Program Instruction VOCR context to choose between the two For example in the text War of 1612 people readily interpret the 1 lowercase L between the 6 and the 2 as the numeral 1 In this mode of operation the following data is returned in addition to that returned for Quick Verification op 0 above data i 3 TRUE if expected characters were all 1st picks data i 4 TRUE if they were all 1st or 2nd picks Operation 2 Robust Verification Full Output This operation is identical to the one above op 1 except more information is returned In addition to the array elements described above the top two picks and their scores are
241. n This VABORT is sent to cancel any pending VPICTURE that might be waiting for a fast digital input interrupt signal The one vision instruction that can wait forever is a VPICTURE that waits for a fast digital input interrupt signal The V IO WAIT system parameter enables this wait mode If the input signal never occurs the VPICTURE request will never complete VABORT however may be used to cancel the request In multitasking applications one task may issue a VABORT to abort a vision instruction that another task is waiting on thereby letting that task resume Only pending or executing vision instructions associated with the given task number are aborted However if the given task number is 2 all pending or executing vision instructions are aborted Example Abort executing or pending vision instructions issued from task 0 VABORT 0 24 AdeptVision Reference Guide Rev A Syntax VADD cam Function Program Instruction VADD type dmode dest_ibr srcl_ibr src2_ibr Add two binary or grayscale images Parameters cam type dmode dest_ibr srcl_ibr Details Selects the threshold parameters to apply to the result Optional integer value indicating the type of addition 1 for binary default value 2 for average grayscale 3 for summed grayscale Optional real valued expression specifying the display mode for this operator The choices are 1 no draw 0 erase 1 draw solid 2 comp
242. n image for defining a prototype This camera will also be the default camera associated with the prototype see the AdeptVision User s Guide for details on associating cameras with a prototype mode Optional integer interpreted as follows O Anew image is acquired when prototype editing begins Repeated training samples are allowed and a new picture is acquired for each 1 The existing image is used This allows you to preprocess an image before training After one sample training exits 2 Disable options in the training pull down menus value controls which options are disabled If New Example is disabled training proceeds similar to mode 1 3 Change a prototype parameter but do not train new samples The parameter to change is specified in arg The new value for the parameter is specified in value Adept Vision Reference Guide Rev A 267 VTRAIN 268 arg prototype ibr shape Program Instruction If mode 3 a prototype parameter is being changed This parameter identifies the parameter to change value supplies the value arg Parameter Range allowed for value 1 Verify percent 0 to 100 2 Effort level 0 to 4 3 Min area in pixels 1 to proto area 1 4 Max area in pixels proto area 1 to image size 5 X constraint in mm 6 Y constraint in mm 7 Rotational constraint 8 Virtual camera bit mask hex 1 to F 9 Virtual camera bit mask hex 10 to F
243. n Reference Guide Rev A Program Instruction VSHOW Syntax VSHOW mode proto_name trans_var grip edge_num Function Display a vision model ObjectFinder prototype or subprototype and make information about it available through the VFEATURE real valued function Usage Considerations The VISION switch must be enabled the vision processor must be idle and vision model training must not be active for this instruction to be executed Parameters mode Optional bit field expression indicating which of several possible events should result from this instruction Defaults to 0 see below for details proto_name String variable to be assigned the name of the vision model accessed for get next mode or a string expression representing the name of the vision model to be shown for get specific mode If a subprototype is desired this parameter must have the form namel name2 where namel is the name of the prototype containing the subprototype name2 trans_var Optional transformation variable to be assigned the location of the prototype in the Vision display window This parameter is not accepted if a subprototype is being accessed grip Optional real valued expression that specifies the grip position to be displayed along with the model The default value is 0 which means no grip positions are shown The value 1 causes all defined grip positions to be displayed A value in the range 1 to 4 displ
244. n edge point was found otherwise FALSE 124 Adept Vision Reference Guide Rev A Program Instruction VFIND POINT data i 1 TRUE if any part of the search window falls off the Vision display window data i 2 data i 3 X Y coordinates of the edge point found in millimeters ibr Integer value specifying the image buffer region within which to search for a point Image buffer regions specify both a size and a frame store see the description of VDEF AOD i Optional array index that identifies the first element to be defined in data The default is 0 If a multiple dimension array is specified only the right most index is incremented as the values are assigned xc yc Real valued expressions specifying the center of the search window in millimeters The values for both xc and yc must be in the range 1000 to 1000 image pixels after they are converted from millimeters length Real valued expression specifying the length of the search window in millimeters The value for length must be in the range 1 to 1000 image pixels after it is converted from millimeters width Real valued expression specifying the width of the search window in millimeters The value for length must be in the range 1 to 1000 image pixels after it is converted from millimeters angle Optional real valued expressions specifying the orientation of the search window in degrees The default is 0 However angles of 90 and 270 degr
245. n memory The image is visible in VDISPLAY mode 1 The histogram is computed by simply reading all the pixels in the image defined by ibr and counting the number of pixels at each intensity level These counts are placed in the array if that optional parameter was specified The image memory read may be any valid image buffer region The region must contain a valid picture Otherwise an error results See the VSELECT program instruction Image histograms are useful for determining the gain V GAIN and offset V OFFSET for the incoming camera data Histograms also help determine the thresholds for binary image processing 146 Adept Vision Reference Guide Rev A Monitor Command and Program Instruction VHISTOGRAM When the histogram is displayed in the Vision display window the range of intensity is represented by the horizontal axis and the number of pixels per intensity is represented along the vertical axis The vertical axis is scaled to accommodate the highest peak That axis is not labeled because usually only the relative magnitudes are important The following text is displayed to the side of the histogram Min aaa Max bbb where aaa is the minimum intensity with a nonzero count of pixels and bbb is the maximum intensity with a nonzero count Note that the VWWINDOWI program instruction can return the same information as VHISTOGRAM VWINDOWI does not display the histogram but it can be executed when the vis
246. n syntax or explicitly include position information Parameters mode Optional argument that must currently be set to 1 The default value is 1 dx dy Optional real values variables or expressions interpreted as millimeter offsets in the X Y Cartesian plane The default value is 0 angle Optional real value variable or expression interpreted as an angular rotation The default value is 0 scale Optional real value variable or expression interpreted as an integer that scales the dimensions of the tool The default is 1 Details The primary use of VDEF TRANS is to place inspection tools relative to a frame or object previously located by the vision system All vision tools regardless of syntax used placed after a VDEF TRANS instruction has been issued will have the indicated transformation applied before the tool is placed Issuing a VDEF TRANS instruction with no arguments defaults of dx dy and angle 0 will cancel the vision transformation Examples Rotate a ruler tool perpendicular to the angle returned by a line finder ibr_rect has been defined for rectangular tools VFIND LINE 1 data ibr_rect Find the line ang data 4 90 Angle for perpen rotation Define the transformation Adept Vision Reference Guide Rev A 73 VDEETRANS Program Instruction V p Place V V EF TRANS ang the ruler ULERI 1 rdata ibr_rect EF TRANS Cancel transformation ORO Y Place an inspection wi
247. n that specifies the processing mode The mode is optional and defaults to 1 Initiates processing of a new camera image Initiates reprocessing of previous camera image Grabs an image and holds it for future processing Quick frame grab Nr OF The above list should actually be listed in reverse order because a VPICTURE in mode 1 is effectively the same as a VPICTURE in mode 2 followed by another in mode 0 Mode 2 acquires an image as does mode 1 and mode 0 processes the image Mode 1 does both Optional real valued expression that specifies the maximum number of objects the vision system should try to recognize This limit is applicable only in VPICTURE modes 1 and 0 See below for details 1 Locate as many as possible the default 0 Locate none V RECOGNITION effectively disabled 1 Locate only one object 2 Locate at most two objects etc VPICTURE modes 1 and 0 initiate processing When this happens the vision system collects the camera data grabs a frame analyzes it to identify objects and displays the results in the Vision display window in the manner dictated by the current VDISPLAY mode The computed part information is then stored in the vision system queue and is accessible with the VLOCATE program instruction and the VFEATURE real valued function Also region boundary descriptions are available with the VEDGE INFO instruction Adept Vision Reference Guide Rev A 195 VPICTURE Monitor Comm
248. nal real valued expression specifying the display mode for this operator The choices are 1 no draw 0 erase 1 draw solid 2 complement 3 draw dashed and 4 complement dashed The default is 1 draw solid dest_ibr Optional integer value specifying the image buffer region to receive the edge image Image buffer regions specify both a size and a frame store see the description of VDEF AOI If this parameter is not specified the edge image will overwrite the source image src_ibr src_ibr Integer value specifying the image buffer region from which to extract edges The image buffer regions AOI must have been defined with a VDEF AOI instruction Details This instruction computes edges from the grayscale values in an image and thresholds the edges Only the binary image associated with dest_ibr is affected If dest_ibr is not specified src_ibr is used as the destination That is the edge operation is performed in place The effect of a VEDGE operation is visible in VDISPLAY mode 2 It does not affect the associated grayscale image The type parameter to VEDGE determines the method of edge extraction The cross gradient operator is faster but the Sobel is a bit more robust See the entry for the system parameter V EDGE TYPE on page 94 for a detailed description of the cross gradient and Sobel methods The smaller the area of the image to be processed the faster the operation executes The cr
249. ncreases as the width of the operator increases And the operator is faster when rotated near 90 or 270 degrees than it is when rotated near 0 or 180 degrees The cost of filtering depends on the number of iterations performed After a line has been fit that is after the final fit if edge filtering is enabled the edge points used to fit the line are compared with the fit line and the distance from the fit line to the most distant edge point is returned data i 6 This distance is in pixels If this value is more than one or so the line edge in the image is very rough or some extraneous edges were detected and used in the line fitting If edge filtering is enabled data i 7 and data i 8 more specifically indicate the most distant edge points from the dark and bright sides of the line respectively Data item data i 6 is the maximum of data i 7 and data i 8 Example 122 The following program segment finds a corner point with high precision where the corner is formed by two linear edges VFIND LINE is used twice one for each line Then the program line_line also shown below is called to compute the point where the two lines intersect A similar program is shown in the programming example in AdeptVision User s Guide cam 1 Define virtual camera VPICTURE cam 2 Quick frame grab PARAMETER V MAX SD cam 1 5 Filter beyond 1 5 Std Devs PARAMETER V MAX PIXEL VAR cam 1
250. nd its length is computed The computation is fast The information is available from the VFEATURE function after a VLOCATE instruction has succeeded This information is also available for holes To get information for holes you should enable V HOLES disable V DISJOINT and do VLOCATEs in get hole mode Example 192 The following program segment computes a roundness factor for each region in the scene based on the region perimeter and area The roundness factor will be close to 1 0 for circles and less than 1 0 for other shapes This may be used in place of prototype recognition to recognize circles ENABLE V BOUNDARIES V PERIMETER Necessary switches DISABLE V SUBTRACT HOLE If this is enabled VFEATURE 40 should be added to the area vf area 10 VFEATURE index for area pixels vf perimeter 41 VFEATURE index for perimeter mm VGETCAL vb Need items in calibration array Define indexes for x scale 1 Pixel scale mm pixels y scale 16 scale SOR vb x scale vb y scale 2 VPICTURE 0 Take picture with virt cam l no recog VLOCATE Snam Locate anything in the image Adept Vision Reference Guide Rev A System Switch V PERIMETER WHILE VFEATURE 1 DO If a region was found area VFEATURE vf area scale Raw pixel area gt mm square perim VFEATURE vf perimeter
251. ndow relative to a found blob ENABLE V CENTROID Get centroid information VDEF AOI win aoi 1 10 40 Postion of window relative 7 to bLob VPICTURE 1 VLOCATE 1 2 Locate a blob IF VFEATURE 1 THEN Make sure a blob was found cx VFEATURE 42 cy VFEATURE 43 ang VFEATURE 48 VDEF TRANS cx cy ang VWINDOWI wdata win aoi END VDEF TRANS Cancel transformation Related Keywords VDEF AOI program instruction VGET AOI program instruction VGET TRANS program instruction 74 Adept Vision Reference Guide Rev A Syntax Monitor Command and Program Instruction VDELETE VDELETE model_name Function Delete a specified prototype subprototype Optical Character Recognition OCR font or correlation template in the vision system Usage Considerations The VISION switch must be enabled and the vision processor must be idle for this command or instruction to be executed Parameters model_name Details Name of the prototype subprototype OCR font or correlation template to be deleted If a subprototype is being deleted the model name has the form namel name2 where namel and name2 are the names of the prototype and its subprototype respectively If a font is being deleted the model name has the form FONT_n Ms FE where n is the number of the font in the range 1 to 99
252. ne a bounding box for the region VLOCATEd or the prototype last VSHOWed The bounding box is an orthogonal rectangle that encloses the region with the four corners Min_X Min_Y Min_X Max_Y Max_X Max_Y Max_X Min_Y One application of a bounding box definition is for finding clear space for placing things on a conveyor belt or a pallet For a region the bounding box coordinates are in the reference frame of the camera For a prototype following a VSHOW instruction the coordinates are with respect to the reference frame of the prototype ITEM 17 Following VLOCATE this is the number of holes detected in the region Following VSHOW this is the number of holes in the prototype ITEM 18 This is the elapsed time in seconds spent to acquire and analyze the image and to perform recognition Since it is elapsed time it includes the time spent to update the display and to respond to any commands such as VSTATUS and VQUEUE The detailed procedure for keeping time is as follows AdeptVision Reference Guide Rev A 99 VFEATURE Real Valued Function 100 Timekeeping starts when the VPICTURE or VWINDOW command is received A time record is kept for each object recognized or region analyzed and not recognized in the image As each object is entered in the queue making it ready to be VLOCATEd the elapsed time is noted Timekeeping is reset as the search begins for the next object in the image Consequently the elapsed time associated
253. ne of pixels to be processed 296 Appendix A AdeptVision Quick Reference V FIT ARCS camera Enable or disable the fitting of circular arcs when performing boundary analysis V GAIN camera Set the gain for the incoming video camera signal VGAPS data i proto_name edge_num Find the unverified gaps in a match with a prototype or subprototype VGET AOI array i aoi Return the definition of an area of interest VGETCAL cam scalers i pmm to pix j k l pix to pmm 1 m to cam Ask the system to fill in arrays with the previously defined vision calibration data for a given virtual camera VGETPIC cam type s_rate s_mode Spic r c shape x0 y0 dx dy Read all or part of an image into a string array VGET TRANS array i Return the value of the current vision transformation VHISTOGRAM dmode array index ibr Compute the histogram for a grayscale frame store V HOLES camera Enable or disable the accounting of interior features in all objects V IO WAIT camera Enable the synchronization of taking pictures VPICTUREs with an external event that triggers the fast digital input interrupt line VISION Enable the entire vision system V LAST COL camera Set the number of the last column of pixels to be processed V LAST LINE camera Set the number of the last line of pixels to be processed V LAST VER DIST camera Enable an extra verification of prototype to image matches and speci
254. ng the picture to the frame store If zoom is N every pixel is blown up to an NxN square The zoom factor must be in the range 1 to 100 The default is 1 NOTE The parentheses in the instruction syntax can be omitted if all three of the above parameters are omitted pic Array of picture data including a header string The data in this array is assumed to have been created by the VGETPIC program instruction came Row and column indexes into the array pic indicating where the header string and picture data are The defaults are 0 for both r and c The header string is in the element pic r c The picture data starts at element pic r 1 c 1 x0 y0 Optional real valued expressions specifying the coordinate of the lower left corner of the image where the picture is to be stored If not specified the starting coordinate stored in the header string is used x0 and yo are in pixel units and must be in the image The maximum value is the size of the virtual frame buffer the image data is being written to Virtual frame buffer sizes are configured 206 AdeptVision Reference Guide Rev A Program Instruction VPUTPIC with the DEVICE instruction See the AdeptVision User s Guide for details Details This instruction provides a means for replacing picture data extracted using the VGETPIC instruction Normally this is used for reviewing an old picture However for diagnostic purposes saved
255. ningful operation It doubles the intensity of the image buffer region and puts the result in dest_ibr Example 26 Add two image buffer regions and place the result in the first image buffer region aoi2 2000 AOI number 2 cx 4 5 Center of AOI cy 8 A size 5 Length of side of AOI rect 1 Use shape definition 1 ang 0 VDEF AOI aoi2 rect cx cy size size ang dest aoi2 11 AOI 2 virt frame buffer 11 AdeptVision Reference Guide Rev A Program Instruction srcl aoi2 11 AOI 2 virt frame buffer 1 src2 ao12 12 AOI 2 virt frame buffer 1 VADD dest srcl src2 Related Keywords VCOPY program instruction VDEF AOI program instruction VEDGE program instruction VGET AOI program instruction VSUBTRACT program instruction VTHRESHOLD program instruction Adept Vision Reference Guide Rev A VADD 27 VAUTOTHR Monitor Command and Program Instruction Syntax VAUTOTHR dmode start end array index ibr Function Determine good thresholds for binary image processing based on the gradients in a grayscale frame store Parameters dmode start end array index ibr Details Optional real valued expression specifying the display mode 1 to display the histogram and average contrast curve or 0 for no display The default is 1 display Optional integer values specifying the range of intensities in which to limit the search
256. nrotated rectangle Image buffer regions specify both a size and a virtual frame buffer see the description of VDEF AOI NOTE There is no optional shape specification it must be an image buffer region ibr 114 Adept Vision Reference Guide Rev A Program Instruction VFINDER Details This instruction performs ObjectFinder recognition using the planning associated with the given virtual camera As shown below several of the same switches and parameters apply as for prototype recognition Input parameters using virtual cameras NOTE All switches and parameters should use the default settings except as noted here The following settings are required PARAMETER V MIN AREA vc 4 PARAMETER V MIN HOLE AREA vc 4 The following settings are suggested V MAX PIXEL VAR vc V EDGE STRENGTH ve V MAX TIME vc 2 0 V MAX VER DIST vc 5 Same as for models trained Same as for models trained assuming that the F stop or illumination has not changed significantly from when the model was trained Used if the times array is omitted 0 The following switches are available for enabling visual feedback V SHOW FBNDS vc V SHOW FEATS vc V SHOW VERIFY ve V SHOW RECOG vc Similar to VSHOW BOUNDS but for ObjectFinder Shows final features used for ObjectFinder Same as for prototype recognition The lines and arcs of the prototype are compared to the edges in the image and they are drawn
257. nt of the object transformation RZ This is the r component of the object transformation Instance ID Number of the object Since an image may contain several objects that match the same prototype this number is provided 208 AdeptVision Reference Guide Rev A Monitor Command VQUEUE to allow a program to unambiguously reference a specific instance The first object encountered after a VPICTURE or VWINDOW instruction is assigned an ID of 1 and each subsequent object is assigned the next sequential number Camera This is the virtual camera number of the object Example VQUEUE 4 Name Verify Area X Y RZ Instance Camera Percent LD FLANGE 99 9 2000 10x1 123 0 30 00 1 4 CASTING 100 0 1234 101 0 49 2 170 14 2 4 0 0 165432 765 3 30 9 0 00 3 4 Related Keyword VQUEUE real valued function Adept Vision Reference Guide Rev A 209 VQUEUE Real Valued Function Syntax VQUEUE camera Sname Function Return the number of objects in the vision system queue Usage Considerations Vision model training must not be active for this operation to execute Parameters camera Optional real valued expression that specifies the virtual camera number The default is 0 meaning all cameras name Optional string expression that specifies the name of a prototype Details If a prototype name is specified VQUEUE returns the number of them in the queue for the given virtual camera If the virtual camera is not specifi
258. nterrelated via parent child and sibling relationships A hole in a region is a child of the outer region and the outer region is a parent of the inner region Two holes in a region are mutual siblings The vision system completes connectivity analysis before proceeding on to the other processing steps Chain Encode Perimeters Processing proceeds with this step after the outermost region the region whose parent is the background closes off This step produces a chain encoding representation of the region perimeter Using the stored run length encodings with their region associations provided by connectivity analysis in step 1 a string of right up left and down moves are produced that will traverse the region perimeter Fit Primitive Edges to Chains This step fits straight line segments to the chain encodings in order to produce a more succinct representation of the region perimeter The algorithm used is efficient and the edges produced are an accurate approximation of the chain encoding In this manual these are called the primitive edges A benefit of the algorithm is that the jagged quantized edges represented by the chain encoding are smoothed After this step the chain encoding data are discarded Fit Lines and Arcs to Edges This processing step fits straight line and circular arc segments to the edges produced in step 3 This step is relatively expensive in processor time but the expense is well justified First of all
259. ntical to predefined convolution number 1 3x3 average flat FOR i 1 TO 7 Zero the entire 7x7 matrix FOR j 1 TO 7 ve i j 0 END END Define the nonzero elements Se FOR i 3 TO 5 FOR j 3 TO 5 ve i j 14 END END VDEF CONVOLVE 17 vc 1 1 Define the convolution as 17 Related Keyword VCONVOLVE program instruction Adept Vision Reference Guide Rev A 61 VDEEFONT Program Instruction Syntax VDEF FONT op font_num chars height black_chars Function Define replace or modify an Optical Character Recognition OCR font Usage Considerations The VISION switch must be enabled the vision processor must be idle and vision model training must not be active for this instruction to be executed Parameters op Optional real valued expression defining the desired operation O define or replace a font 1 modify an existing font The default is 0 font_num Real valued expression defining the number of the font being defined replaced or modified The number must in the range 1 to 99 inclusive chars String expression containing all of the characters in the font in any order See below for more information height Optional real valued expression defining the typical height in pixels of the tallest character in the font The value must be in the range 6 to 63 pixels Note if characters in the font are wider than they are tall the maximum width should be specifi
260. ntrol pendant to move the robot and interact with motion control programs V Operating System User s Guide Details on the V operating system V Operating System Reference Guide Detailed descriptions of the V monitor commands AdeptVision Reference Guide Rev A Chapter 1 AdeptVision Keyword Summary Table 1 1 Related Publications Continued Manual Material Covered V Language User s Guide V is a complete high level language as well as an operating system This manual covers programming principles for creating V programs V Language Reference Guide Detailed descriptions of the keywords in the V language User s guides for any AIM software that Details on using AIM applications such as MotionWare or may be installed on your system VisionWare AIM software reference guides Details on the structure of AIM software and AIM applications AdeptVision Keyword Summary Table 1 2 summarizes the keywords added with the AdeptVision VXL option Appendix A is a quick reference guide to the AdeptVision keywords listing their arguments and basic functions Table 1 2 AdeptVision Keyword Summary Keyword Description V 2ND MOMENTS Enable computation of the best fit ellipse for each region in the image V 2ND THRESH Set a second threshold for use during binary image processing VABORT Abort any active vision proces
261. number virtual frame buffer physical frame store When VVP 0 the currently selected virtual frame buffer is used When NNN 0 1 is assumed full frame NOTE Instructions requiring two areas of interest that are the same size will reduce the larger area of interest about its center Example fs2 2 Physical frame store 2 fsl 1 Physical frame store 1 virt_fr 2 10 Virtual frame store 2 Define two areas of interest aoi_up 8 1000 Area of interest 8 aoi_dwn 12 1000 Area of interest 12 VDEF AOI aoi_up 1 10 15 20 10 90 VDEF AOI aoi_dwn 1 10 5 20 10 90 54 The special value VVP 009 refers to the display frame store This frame store contains the image displayed on the monitor and can be referenced only by a VCOPY instruction AdeptVision Reference Guide Rev A Program Instruction VDEF AOI Define three image buffer regions src_ibrl aoi_uptvirt_frt fsl AOI 8 virtual frame 21 src_ibr2 aoi_dwn virt_fr fs2 AOI 12 virtual frame 22 dest_ibr aoi_up virt_fr fs2 AOI 8 virtual frame 22 Add AOI 8 in virtual frame 21 src_ibrl 8021 to AOI 12 in virtual frame 22 src_ibr2 12022 Place the result in AOI 8 in virtual frame 22 dest_ibr 8022 VADD 1 dest_ibr src_ibrl src_ibr2 Related Keywords VADD program instruction VAUTOTHR program instruction VCONVOLVE program ins
262. o a single bounded area but inaccuracies accumulate if more characters are merged together As a general rule VOCR splits two characters very well splits three merged characters moderately well and splits four or more characters poorly Example The following V program uses the VOCR instruction to perform character recognition or verification depending on the value of the variable op The results are displayed in the Monitor window a box is drawn around the text in the Vision display window and if op is 2 or 3 a dot is drawn at the center of each character found NOTE This example program assumes that the OCR font has already been defined and trained PROGRAM test ocr ABSTRACTExample program showing basic OCR capabilities AUTO ang cam cc cnt cx Cy dx dy dm font op AUTO Sans Sexp LOCAL xx data locs Change these values to suit your needs font 1 Font number op 2 VOCR operation 0 to 3 Sexp ABCDEFG Expected text required if op lt gt 3 cx 256 Center of the box cy 242 dx 350 Width and height of the box dy 50 ang 0 Orientation of the box cam 1 Camera number dm 1 Display mode VDISPLAY cam 2 1 VPICTURE cam 2 WHILE TRUE DO VWINDOWI xx cx cy dx dy ang Draw window VDISPLAY cam 0 1 Show live video TYPE Position the text in the window shown S PROMPT Press Enter when ready Sans
263. o read parameters from type Real valued expression indicating the type of planning 0 Basic initial planning 1 Quick initial planning for multi instance training dmode Not used Sfmods An array of foreground models Each list must contain names of valid trained models starting with index 0 and terminating with a null string as the last name Sbmods An array of background models Each list must contain names of valid trained models starting with index 0 and terminating with a null string as the last name Details Foreground models are those for which you would like the Finder to return locations Background models are those that may be present but which you are not interested in locating Specifying these when applicable will help planning tell which features of the foreground models are best for finding proposals that are not ambiguous with respect to other similar looking objects in the scene Description of planning types Type 0 planning takes the lists of models and does all the preparation work needed for recognition ahead of time to allow for the fastest possible recognition times Planning is usually fairly fast but can take many seconds to compute and many Kb to store from V program memory if the total number of model pairs is large Type 1 planning is faster than Type 1 planning but causes recognition to be slower However you may achieve more reliable results using Type 2 It does this by several means
264. o the side of the window containing the initial search point The search lines are spaced one pixel apart If the effort level is 100 all search lines are searched for edge points If the effort level is 50 only every other search line is searched for edges The effort level is basically the percentage of all potential search lines that will be searched for edge points For very high accuracy in locating a corner formed by two linear edges the VFIND LINE instruction should be used See the program example given with VFIND LINE If one or both of the edges are circular arcs the VFIND ARC instruction may be used in a similar manner If part of the operator falls off the Vision display window that is if part of the search window must be clipped to the virtual frame store data i 1 notes it and the vision system still searches for edge points The execution time for VFIND POINT is proportional to the length of the search window Also the execution time increases as the width of the window increases and as the effort level increases The operator is faster when rotated near 90 or 270 degrees than it is when rotated near 0 or 180 degrees Example The following instruction sequence uses the V EDGE STRENGTH system parameter for camera 3 to find the right most edge point in a search rectangle The search is from the light side pos 1 of a rectangle located at 100 67 33 The search window is 40 millimeters high and 25 millim
265. occurs and overall image quality is best However since the light is strobed immediately after a field transfer the frame grabber must wait one field time before acquiring a frame because the first field will not have been exposed to the strobe light This adds a constant 16 7 milliseconds to the execution time of a VPICTURE Consequently using this method the fastest ping pong processing rate is 20 frames per second instead of 30 Another problem with firing the strobe at a fixed time is the variable time between the VPICTURE request or the external event triggering the VPICTURE and the firing of the strobe The worst case delay is 16 7 milliseconds This worst case delay should be planned for if objects in the scene are moving fast with respect to the camera field of view AdeptVision Reference Guide Rev A System Parameter V SYNC STROBE What is the effect of a worst case delay of 16 7 milliseconds Consider for example a field al view that is 6 4 centimeters wide Then each pixel is 0 1 millimeter wide If objects in the field of view are moving at the high speed of 1 meter per Seon a worst case delay of 16 7 milliseconds corresponds to a shift of 167 pixels Since the frame store is 640 pixels wide a 167 pixel shift could shift part of the object being analyzed out of the field of view When you are using a strobe light synchronously you should perform the above calculations for your application and adjust the setup accordin
266. ode 0 or 1 or VWINDOW is executed The Vision display window must have a graphical display mode in effect This is an array of switches one for each virtual camera See the AdeptVision User s Guide for details on setting switches Details When V SHOW EDGES is enabled edges are displayed when certain vision operations are performed There are two groups of operations that display edges when V SHOW EDGES is enabled The first group consists of the finders VFIND LINE VFIND ARC and VFIND POINT When V SHOW EDGES is enabled the edge points that are found in the area of interest of a finder are displayed in the Vision display window VFIND LINE and VFIND ARC display in white the edge points that were used to fit a line or an arc respectively Edge points that were filtered out are displayed in gray VFIND POINT displays in white all the edge points it found VFIND POINT does not filter out edge points The display of edge points by the finders is very costly in terms of processing time so VSHOW EDGES should be enabled only for investigative purposes The other operations that display edges when V SHOW EDGES is enabled involve boundary analysis Both VPICTURE in modes 0 and 1 and VWINDOW perform boundary analysis When V SHOW EDGES is enabled the connected sequence of primitive straight lines that bound the regions are displayed in the Vision display window overlaying the displayed regions The lines are joined by corners whi
267. ode dmode data index ibr Function Extract image information from within a rotatable rectangular window Usage Considerations All bounded regions in the image buffer region are considered one blob The statistics returned are for all bounded regions regardless of whether they are disjoint The perimeter calculations for this instruction are less accurate than for VWINDOWT and are available only with the AdeptVision Enhanced VXL Interface option Parameters cam Optional real valued expression that specifies a virtual camera number The system parameters V THRESHOLD and V 2ND THRESH and the system switch V BACKLIGHT for this virtual camera are used during processing The default is 1 mode Optional real valued expression that specifies what image statistics are to be computed and returned in the data array as follows 1 Default Computes the area bounding box and centroid Time 17ms 2 For EVI board option only Computes all of the above plus perimeter Time 25ms dmode Optional real valued expression that specifies the display mode to use when displaying the border of the window 1 No draw O Erase 1 Default Draw solid 2 Complement 3 Draw dashed 4 Complement dashed NOTE The parentheses in the instruction syntax can be omitted if all three of the above parameters are omitted data Variable name specifying an array into which the image information is to be placed The amount of information depends on the type
268. odel is specified If a specific OCR font is referenced data i 0 Number of characters in the font data i 1 Height of the font in pixels data i 2 Font color TRUE black FALSE white data i 2 1 Number of trained instances of the first character in Schars data i 2 n Number of trained instances of the nth character in Schars If a specific correlation template is referenced 234 AdeptVision Reference Guide Rev A Program Instruction VSHOW MODEL data i 0 Width of the template in pixels data i 1 Height of the template in pixels If all fonts or templates are referenced data i 0 Number of fonts or templates defined data i 1 Number of first font or template defined data i n Number of nth font or template defined i Optional array index that identifies the first element to be defined in aata The default is 0 If a multiple dimension array is specified only the right most index is incremented as the values are assigned model_num String variable specifying a font all fonts a template or all templates for display if mode 0 and return of information The string naming a font has the form FONT_n where n is the number in the range 1 to 99 of a single font or FONT_0 specifies all fonts Similarly the name of a template has the form TMPL_n where n is the number in the range 1 to 99 of a single template TMPL_0 specifies all templ
269. olid how_many Optional real valued expression that specifies the maximum number of objects the vision system should try to recognize in the window See the description of VPICTURE for a detailed explanation Locate as many objects as possible the default Locate none V RECOGNITION effectively disabled Locate only one object Locate at most two objects Locate at most n objects SNR OF 284 AdeptVision Reference Guide Rev A Program Instruction VWINDOW NOTE The parentheses in the instruction syntax can be omitted if all four of the above parameters are omitted ibr Integer value specifying the image buffer region for the window Image buffer regions specify both a size and a frame store see the description of VDEF AOI The image buffer regions AOI must specify a rectangular shape shape Optional real valued expression indicating the shape of the window Currently the only choice is 1 for rectangular cx cy Real valued expressions specifying the center coordinate of the rectangular window in millimeters width Real valued expression specifying the width of the window in millimeters height Real valued expression specifying the height of the window in millimeters ang Optional real valued expression specifying the orientation of the window in degrees The default is 0 degrees Details VWINDOW performs image processing within an area of interest called a window A window is a rectangle of any size and or
270. omputing their average distance to the transformed center point Adept Vision Reference Guide Rev A 311 Appendix C Calibration Arrays 312 PROGRAM tr point trans pt tptI l ABSTRACT Transform a 2 D point given a 3x3 homogeneous transformation INPUT PARAM trans 3x3 real array containing a 2 D homogeneous A transform with pixel to mm scaling and compensation for perspective ptt 3 element real array containing the X Y Z coordinates of the point to be transformed 5 the Z value should be 1 unless you are i scaling i OUTPUT PARAM tpt 3 element real array containing the X Y Z 7 coordinates of the transformed point normalized so that the Z value is 1 Copyright c 1992 by Adept Technology Inc AUTO row col sum FOR row 0 TO 2 Compute transformed point sum 0 FOR col 0 TO 2 sum sum trans row col pt col END tpt row sum END tpt 0 tpt 0 tpt 2 Normalize the result tpt 1 tpt 1 tpt 2 tpt 2 1 END As an example of how to use the routine tr point the following instructions apply perspective correction to the centroid of a blob They would be executed after the blob has been successfully VLOCATEd pt 0 VFEATURE 2 X coordinate in simple mm space pt 1 VFEATURE 3 Y y MA pt 2 1 CALL tr point mm to pmm pt exact_pt After these instructions are executed the elements exact_pt 0 and exa
271. on array is specified only the right most index is incremented as the values are assigned Optional string expression that specifies the name of the prototype or subprototype for which gap information is requested If a subprototype is specified the string must have the form namel name2 where namel is the name of the prototype and name2 is the name of the prototype s subprototype The default if proto_name is not specified is the prototype last VLOCATEd AdeptVision Reference Guide Rev A Program Instruction VGAPS edge_num Optional real valued expression that specifies the number of the prototype or subprototype edge for which information is requested The default is 0 which requests a list of all the edge numbers that have gaps in them If the value is nonzero gap information about a specific edge is requested The edge number is relative to the prototype s edge numbering unless proto_name specifies a subprototype Then the edge_num is relative to the subprototype s edge numbering Details When a prototype object is recognized in the image each edge of the prototype is compared with the edges in the image to verify the prototype match When the V LAST VER DIST system parameter is nonzero the portions of prototype edges that are not verified are remembered These unverified portions are called gaps Gaps could be caused by cutouts flashing occluding debris or by dimensions that are out o
272. on the strobe device must be met for successful strobe operation 1 The combination of the flash latency from trigger input and the flash duration must not exceed 100 microseconds AdeptVision Reference Guide Rev A System Switch V STROBE 2 The polarity required by the strobe device must match the strobe polarity that is specified in the camera model All the default camera models are set for active high This can be changed with the DEVICE instruction 3 The duration of the strobe signal 120 microseconds must be acceptable to the strobe device A VPICTURE instruction in wait mode normally waits for the acquisition of the camera image to finish before the next instruction in the V program starts to execute However if the VSTROBE switch is enabled the VPICTURE completes after the strobe signal letting the V program resume earlier than usual That is done because the strobe light is assumed to be sufficiently bright to freeze the image even if the camera is mounted on a robot and the next instruction is a robot motion If this assumption is not valid for the application the application program should execute a VWAIT instruction after the VPICTURE and prior to any other instructions that may cause the image to change Simultaneous with the firing of the strobe light the positions of all the encoders connected to the Adept system are automatically latched That is the positions are read and stored intern
273. onent RX is always 0 and RY is always 180 RZ is the orientation of the object in the horizontal plane the image plane The setting of the V CENTROID switch may affect the location components If the region is unknown that is the object name is and the V CENTROID switch is enabled the X Y location is for the region centroid and RZ is 0 If the V CENTROID switch is disabled the location is the center of the bounding box for the unknown region The setting of the V CENTROID switch has the same effect on the location components after a VLOCATE in find hole mode However after a VSHOW instruction in get hole mode the X Y location is always the centroid of the prototype After VLOCATE the coordinates are in the reference frame of the camera After a VSHOW instruction the coordinates are in the reference frame of the prototype origin of the coordinate system is the prototype centroid Note that VSHOW draws the prototype in the Vision display window so that the center of its bounding box is at the center of the window ITEM 8 This VFEATURE item is the belt encoder value at the time the strobe light fired see the VSTROBE system switch This value is also available from the V DEVICE real valued function AdeptVision Reference Guide Rev A Real Valued Function VFEATURE ITEM 9 Verify percentage is a recognition certainty value The verify percentage is 0 if the region VLOCATEd is unknown Otherwise it is the percentage of t
274. ormations of binary or edge images The VDEF MORPH instruction defines the operation to perform The VMORPH instruction actually performs the operation Morphological operations may be used to eliminate small holes and gaps from the image by dilating and then eroding or vice versa to thin edges isolate certain features such as straight lines etc Multiple operations are often performed in sequence Morphological operation types 1 and 2 are predefined as erosion and dilation Types 3 through 8 are undefined but reserved to be additional built in operations in the future Types 9 through 16 can be defined by the user Type 9 is actually predefined to be the game of Life popularized years ago in Scientific American but it can be redefined by the user Binary morphological operations are applied to every 3x3 pixel neighborhood in the binary image Based on the binary pixel values in a neighborhood and the operation to be performed the center binary pixel value may be changed The parameter array which Adept Vision Reference Guide Rev A 69 VDEFMORPH Monitor Command and Program Instruction defines the operation actually defines a look up table To index into the table the vision system constructs the nine bits of the index from a 3x3 pixel neighborhood The binary values of the nine pixels in a neighborhood about the center pixel 5 are as follows 9 8 7 6 5 4 3 2 1 Each of the pixel values 1 through 9 has the value 0
275. osition of the center of the search window in millimeters Real valued expressions specifying the width dx and height dy of the search window in millimeters Not currently used AdeptVision Reference Guide Rev A Program Instruction VCORRELATE Details Normalized Grayscale Correlation NGC is an image to template comparison More than simply an image subtraction normalized correlation normalizes the differences accounting for lighting or contrast changes Additive or multiplicative changes to the image do not affect the correlation results Binary correlation simply compares the black white thresholded image with the binary template To define a template you merely define an orthogonal rectangle in a desired image using the VTRAIN MODEL instruction The pixels within the image are stored in the vision CPU memory More than one template may be defined and they may be stored to disk for later recall see the VSTORE and VLOAD keywords for details The pull down menus on the Vision display window let you list show delete and rename templates Templates are numbered in the range 1 to 99 Some monitor commands and instructions that reference templates use a string name of the form TMPL_n Ml Wy where n is the template number in the range 1 to 99 One of the most important features is the ability to skip rows and or columns when correlating at the most reduced levels of hierarchy Not performing correlation matc
276. oss gradient operator is faster than the Sobel operator Adept Vision Reference Guide Rev A 85 VEDGE Program Instruction Example Compute cross gradient edges for all of frame store 1 using 22 as the edge strength PARAMETER V EDGE STRENGTH 1 22 VEDGE 1 1 1001 Related Keywords VTHRESHOLD program instruction V EDGE TYPE system parameter 86 AdeptVision Reference Guide Rev A Syntax Program Instruction VEDGE INFO VEDGE INFO data i proto_nam edge_num Function Retrieve information about the edges and corners of a prototype or of a region in the image Usage Considerations The V EDGE INFO switch must be enabled before a VPICTURE in modes 0 or 1 or VWINDOW instruction is executed in order to use VEDGE INFO to retrieve information about regions in the image Also if V DISJOINT is true edge information is not available for regions that helped verify the recognition of a prototype Vision model training must be inactive for this instruction to execute Parameters data proto_nam Real array containing the requested edge information data i 0 data i 1 0 line 1 convex arc or 1 concave arc Weight 0 to 100 for prototype edges only data i 2 X coordinate of edge starting corner data i 3 Y data i 4 X coordinate of edge ending corner data i 5 Y data i 6 X coordinate of arc circle center data i 7 Y bl j data i 8 Radius of ar
277. ought if the effort level is 1 If the effort level is 100 as many edge points as possible are sought The effort level affects both execution time and accuracy If part of the operator falls off the Vision display window that is if part of the search window must be clipped to the virtual frame store data i 1 notes it and the vision system still attempts to fit a line However no line is fit if fewer than two edge points are found After a line has been fit the edge points found are compared with the fit line and the distance from the fit line to the most distant edge point is returned data i 6 This distance is in pixels If this value is not close to one the linear edge in the image is very rough or some extraneous edges were detected and used in the line fitting NOTE For highest accuracy when fitting a line to a linear edge the corners at the ends of the linear edge should not fall within the VFIND LINE search window Parts of the other edges that are connected to the corners may be detected by VFIND LINE and used in the least squares fit This shifts the fit line away from the linear edge of interest toward one or both of the neighboring edges VFIND LINE optionally filters out edge points that are far from the fit edge and then refits the edge This is an iterative process that increases the execution time but provides more accurate and consistent results particularly when noise is present in the image Two system para
278. ow specification is the same as that for the VWINDOW instruction which is used to perform object recognition within windows Once VTRAIN has been initiated the mouse is used to interact with the system The user can abort a VTRAIN monitor command at any time by entering Ctrl C See the Adept Vision User s Guide for a description of the training process Exa mple Initiate vision training for a prototype named flange VTRAIN flange Related Keywords VDEEFAOI program instruction VLOCATE program instruction VSTORE program instruction 266 Adept Vision Reference Guide Rev A Program Instruction VTRAIN Syntax VTRAIN cam mode arg prototype ibr value VTRAIN cam mode arg prototype shape cx cy dx dy ang value Function Initiate training of the prototype whose name is specified Usage Considerations This instruction requires that all virtual cameras be idle that is no VPICTURE or VWINDOW operation can be executing Prototype training is sensitive to camera calibration Before training prototypes mount focus and securely fix your cameras and lenses Then calibrate the camera s that will be used to train and recognize the prototypes The same camera s with the same calibration and position used to train a prototype must be used to recognize the prototype Parameters cam Optional integer specifying the virtual camera and associated calibration switches and parameters to use when acquiring a
279. pe 4 windows the standard deviation of the graylevels from their mean is returned This is the most time consuming value to compute However it provides a more reliable measure of graylevel distribution or edginess than the minimum and maximum graylevel values that are returned Adept Vision Reference Guide Rev A 293 VWINDOWI Program Instruction A type 5 window finds omnidirectional edge points in the grayscale image and counts them The edge detection method used is the same as that used by VPICTURE when the V BINARY system switch is disabled However the setting of V BINARY has no effect on this window operation The VEDGE STRENGTH system parameter is the threshold criterion used As V EDGE STRENGTH is reduced more edges are found in the window As it is increased fewer edges are found The edge points in an image may be viewed in VDISPLAY mode 2 after performing a VEDGE operation of type 1 for cross gradient Rectangular nonrotated windows that is ones with sides parallel to the sides of the image are processed faster than circular windows or rotated rectangular windows Examples Using the system switches and parameters for virtual camera 14 during the processing ask for all the information possible about a type 4 rectangular window centered at the point 230 340 20 millimeters wide and 10 millimeters high with a 15 degree rotation Draw the window boundary in the Vision display window dmode 1 VWINDOW
280. pe bit field 17 to 32 32 First edge number 33 Last edge number 34 X constraint of prototype mm defined during training 35 Y constraint of prototype 36 Angular constraint of proto 37 39 Reserved for future use 40 Area of all holes raw pixels 41 Perimeter millimeters 42 Centroid X millimeters 43 Centroid Y millimeters 44 Minimum radius angle degrees 45 Maximum radius angle degrees 46 Minimum radius millimeters 47 Maximum radius millimeters AdeptVision Reference Guide Rev A 97 VFEATURE Real Valued Function 98 Table 2 1 VFEATURE Function Data Continued Index Value Units comment 48 2nd moments major axis degrees 49 Major ellipse radius millimeters 50 Minor ellipse radius millimeters ITEM 1 VFEATURE 1 has the value TRUE when the object or prototype information is valid otherwise it is FALSE The information is valid when a VLOCATE succeeds in locating an object in an image or VSHOW succeeds in getting a prototype from the loaded prototype list The information is invalid when a no wait VLOCATE fails to locate an object or a VSHOW instruction fails to find a loaded prototype An error occurs if the object information is invalid and a VFEATURE is executed with an index other than 1 ITEMS 2 to 7 X Y Z RX RY and RZ are the components of the object or prototype transformation Note that the complete transformation is optionally returned by the VLOCATE and VSHOW instructions The comp
281. pe of finder whereas they will not be taken into account for the other binary finders type 1 and 2 Again the image data is binary or edge data depending on the setting of the V BINARY system switch Fine edge point finders type 2 find edge points with subpixel accuracy The image data used is in the grayscale frame store which is visible in the Vision display window in VDISPLAY mode 1 VFIND POINT is similar to the VFIND LINE instruction The search window is a parallelogram Graphically the operator looks the same except the center guideline is missing As shown below the knot 0 specifying the initial search position is on either one side of the window or the other pos 1 pos 1 Given the search window on the left above the edge finder would return the X Y coordinate of the edge point in the window that is closest to the top of the window In the search window on the right the edge point closest to the bottom of the window would be returned The VFIND POINT operator has a polarity that determines whether a light to dark or dark to light edge is being fit The vision system assumes that the dark side of the edge is on the top and the light side is on the bottom when the tool rotation is 0 For visual reference one half of the search window is drawn in dark blue and the other half is drawn in light blue The initial search point is drawn in yellow The edge point found is drawn in red If the VSHOW EDGES
282. pression specifying the length of the linear ruler in millimeters ang Optional real valued expression indicating the angle of the linear ruler in degrees measured counter clockwise from the X axis using the start of the ruler as the origin The default is 0 horizontal ruler cx cy Real valued expressions specifying the center of the arc ruler s circle in millimeters radius Real valued expression specifying the radius of the arc ruler s circle in millimeters ang0 angn Optional real valued expressions specifying the angular range of the arc ruler in degrees Each value defaults to 0 A full circular arc is performed if the two angles are equal Details Ruler types 1 and 0 operate on binary data These rulers determine locations along a directed line or circular arc of black to white or white to black transitions The starting color returned is the color black 0 white 1 of the pixel nearest to the start of the ruler Each color change along the ruler is returned as a transition point measured in millimeters or degrees from the start of the ruler The transition points returned are accurate within about one pixel Ruler types 2 1 2 and 3 operate on grayscale data These ruler types must be preceded by a VPICTURE in modes 1 0 or 2 that is any mode but 1 future frame grab The setting of the V BINARY system switch has no effect on these rulers These ruler types can operate on either of the two g
283. r every neighborhood the edge magnitude has been computed Edge above and compared with the V EDGE STRENGTH system parameter If the edge magnitude is greater than the threshold the binary edge bit is set to 1 Otherwise it is set to 0 To keep the threshold values comparable for the two edge operators a normalization factor is applied to the result of the Sobel operator before the threshold comparison Therefore the V EDGE STRENGTH threshold used with one edge operator is generally useful for the other However you should always select the threshold specifically for the operator being used Example Perform the cross gradient for all cameras PARAMETER V EDGE TYPE 1 Related Keywords VEDGE program instruction VPICTURE program instruction V BINARY system parameter V EDGE STRENGTH system parameter AdeptVision Reference Guide Rev A 95 VFEATURE Real Valued Function Syntax VFEATURE index Function Return specified information about the object most recently VLOCATEd or the prototype most recently displayed by the VSHOW program instruction Usage Considerations The VFEATURE function refers to the object most recently VLOCATEd or VSHOWed regardless of which program task executed the VLOCATE or VSHOW instruction Consequently for predictable operation only one program should execute VLOCATE and VSHOW instructions The VSHOW monitor command has no effect on the data available with this function
284. r of the guideline center line in the search window in millimeters The values for both xc and yc must be in the range 1000 to 1000 image pixels after being converted from millimeters Real valued expression specifying the length of the guideline length of the search window in millimeters The value for length must be in the range 1 to 1000 image pixels after it is converted from millimeters Real valued expression specifying the width of the search window in millimeters The value for width must be in the range 1 to 1000 image pixels after it is converted from millimeters Optional real valued expressions specifying the angle of the guideline in degrees The default is 0 However angles of 90 and 270 degrees are most efficient VFIND LINE finds a linear edge in a window with subpixel accuracy if a grayscale tool is used First edge points are found and then a line is fit to the edge points using a least squares fit algorithm VFIND LINE types 2 1 and 0 are binary line finders That is lines are fit to binary edge points which are black to white or white to black transitions in the image Adept Vision Reference Guide Rev A 117 VFIND LINE Program Instruction 118 Dynamic binary line finders type 2 use the grayscale frame store This is the data visible in VDISPLAY mode 1 The values of the threshold parameters V THRESHOLD and V 2ND THRESH at the time the finder is executed are use
285. r region VLOCATEd If a prototype the count includes holes If a region the count does not include holes The bounds counted are lines and arcs A solid disk has one bound a rectangular plate has four bounds etc In get hole mode it is the number of bounds in the region for the hole ITEMS 28 and 29 After VSHOW only These are the minimum and maximum areas associated with the prototype The values are assigned to the prototype during training ITEMS 30 and 31 After VSHOW only These two items are bit masks representing the virtual cameras associated with the prototype Both items are 16 bit fields with each bit representing a different virtual camera Item 30 represents virtual cameras 1 through 16 The least significant bit of item 30 represents camera number 1 and the highest bit represents camera number 16 Item 31 represents cameras 17 through 32 with the ordering similar to that of item 30 The prototype is associated with those cameras whose corresponding bits are set to 1 For example the value 6 in item 30 means that cameras 2 and 3 are associated with the prototype B10 B100 6 ITEMS 32 and 33 After VSHOW only After VSHOW of a prototype item 32 is 1 and item 33 is the number of bounds in the prototype s main region not counting holes These items define the range of edge numbers for the main region After VSHOW in get subproto or get hole mode items 32 and 33 contain the range of edge numbers for the
286. r storing objects in the various vision queues one queue for each virtual camera This memory allocation can be changed with the DEVICE instruction See the AdeptVision User s Guide for details The setting of the VCENTROID switch may affect the location value returned by VLOCATE in find any mode If the region located is unknown that is the object name is and the V CENTROID system switch is enabled the location returned is for the region centroid If the VCENTROID switch is disabled the location is the center of the bounding box for the unknown region The setting of the V CENTROID switch has the same effect on the location returned by a VLOCATE in find hole mode The V CENTROID switch does not affect the location returned by VLOCATE in find particular mode In that mode the location is always for the region centroid Exa mple Locate an instance of the prototype named OBJ found in virtual camera 3 waiting if necessary and then assign the object location to cx VLOCATE 3 2 OBJ cx Get the name and location of the next object or region in the queue in wait mode VLOCATE 0 name loc AdeptVision Reference Guide Rev A 163 VLOCATE Program Instruction Return the name and location of the next object in the queue for virtual camera 2 but do not wait if no object is found VLOCATE 2 1 name loc Related Keywords V BOUNDARIES system switch VFEATURE real valued function VFINDER
287. r use of the program instructions VEDGE INFO VDEESUBPROTO VSUBPROTO and VGAPS Examples Display the model named casting and put its location on the display into the transformation cx VSHOW 2 CASTING cx Create in protos a list of the models currently loaded i 0 VSHOW B11000 Sprotos il Get first model WHILE VFEATURE 1 DO i i 1 VSHOW B01000 Sprotos i Get next in the list END Related Keywords VEDGE INFO program instruction VFEATURE real valued function VLOCATE program instruction VSHOW monitor command 228 AdeptVision Reference Guide Rev A System Switch V SHOW BOUNDS Syntax V SHOW BOUNDS camera Function Enable the special display of the lines and arcs fit to the boundaries of regions Usage Considerations A change to this switch takes effect when the next region is displayed in the Vision display window with a graphical display mode in effect This is an array of switches one for each virtual camera See the AdeptVision User s Guide for details on setting switches Details When V SHOW BOUNDS is enabled the connected sequence of lines and arcs that bound regions are displayed in the Vision display window overlaying the displayed regions Each connected pair of lines or arcs is joined at a corner which looks like a little knot The bounds are displayed in a unique color to distinguish them from other graphics If VSHOW BOUNDS V
288. ra Set the minimum area below which the vision system ignores regions V MIN HOLE AREA camera Set the minimum area below which the vision system ignores holes V MIN LEN camera Set the minimum length of features to be used for feature pairs V MIN MAX RADIT camera Enable the feature that for each region in the image finds the two points on the perimeter that are closest to and farthest from the region centroid VMORPH cam type dmode thresh dest_ibr count src_ibr Perform a morphological transform on a binary image frame VOCR cam font_num VOCR cam font_num ang Perform Optical Character Recognition OCR or text verification in a rectangular image window data i ibr datali l shape op dmode Sexpected op dmode Sexpected locs j locs j cx cy AX dy V OFFSET camera Set the offset for the incoming video signal that is program the zero reference for the A D converter V OVERLAPPING camera Determine whether or not objects may be overlapping in the image Adept Vision Reference Guide Rev A V MAX AREA V PERIMETER camera Enable computation of the lengths of region perimeters VPICTURE camera mode how_many Acquire an image into a frame store and or initiate processing wait acq_ibr sel_ibr VPLAN FINDER Sbmods Set up the type of planning used by the Finder when locating models cam dmode Sfmods type VPUTCAL camera sc
289. ram instruction VSELECT program instruction AdeptVision Reference Guide Rev A 207 VQUEUE Monitor Command Syntax VQUEUE camera Function Display object information for any objects queued in the vision system awaiting retrieval by VLOCATE instructions Usage Considerations This command may be used when a program is executing but vision model training must not be active Parameter camera Optional real valued expression that specifies the virtual camera number If this parameter is specified and is nonzero only objects found for that virtual camera are listed Otherwise regions found for all cameras are listed starting with those most recently VPICTUREd or VWINDOWed The parentheses can be omitted if no virtual camera is specified Details After the vision system has processed an object the object information generated by the recognition algorithms is stored in the vision system until a VLOCATE instruction is executed The VOUEUE command displays the following information for each queued object grouped by the virtual camera number they were found in Name The name of the prototype that matches the object If no match is found the object name is Verify Percent Weighted percentage of the prototype boundary that matches the processed image Area The number of camera pixels enclosed by the object s region boundary X This is the X component of the object transformation Y This is the Y compone
290. rameter sets the approximate maximum number of seconds the vision system will spend either analyzing a region and recognizing an object or performing character recognition see the VOCR instruction This limit keeps the vision system from taking too much time analyzing complex unrecognizable images For prototype recognition note that the maximum time is per object not per region Thus for example if V MAX TIME is 2 and three parts touch in a region the vision system could spend up to 6 seconds analyzing that region V MAX TIME is set to 5 when the V and AdeptVision systems are loaded into memory from disk The range is 0 to 999 A value of 0 means that there is no time limit Example Allow 2 5 seconds for the vision system to recognize each object PARAMETER V MAX TIME 2 5 172 AdeptVision Reference Guide Rev A System Parameter V MAX VER DIST Syntax V MAX VER DIST camera Function Set the pixel tolerance for determining boundary coincidence during the verifica tion of prototype to image matches Usage Considerations A change to this parameter takes effect when the next VPICTURE command or instruction or VWINDOW instruction is executed This is an array of parameters one for each virtual camera See the AdeptVision User s Guide for a general description of parameters Details During the processing of object recognition prototype to image matches are proposed and verified To verify a match th
291. rameter specifies the maximum pixel distance from the fit edge beyond which edge points may be fil tered out During boundary analysis this parameter sets the maximum pixel deviation allowed when fitting lines and arcs to region edges Usage Considerations A change to this parameter takes effect when a new region is analyzed or when a VFIND LINE or VFIND ARC instruction is executed The effects of changes to this parameter are visible in the Vision display window only during training or in a special display mode after a VPICTURE VWWINDOW VTRAIN VFIND LINE or VFIND ARC operation This is an array of parameters one for each virtual camera See the AdeptVision User s Guide for details on setting parameters Details This parameter has two different functions depending on the image processing operation being performed The line and arc finders VFIND LINE and VFIND ARC use it as a distance threshold for controlling the filtering of edge points The finders optionally iterate discarding edge points that are too far from the fit line or arc The parameter V MAX PIXEL VAR specifies the maximum distance from the fit edge within which edge points are always preserved See the parameter V MAX SD on page 170 for a more complete description of the filtering algorithm used by the finder As a general guideline when the finders operate in grayscale mode V MAX PIXEL VAR should be set to 1 or larger When operating in binary mode it sho
292. ransformations such as the VCOPY and VMORPH instructions When subtracting binary images the two source image buffer regions are exclusive ORed XOR and the result is stored in dest_ibr The underlying grayscale image is left unmodified Grayscale averaging is defined as follows dest_ibr srcl_ibr src2_ibr 128 2 AdeptVision Reference Guide Rev A 253 VSUBTRACT Program Instruction The 128 normalizes the output so that no difference has the mid gray value 64 When the src1_ibr pixel value is greater than the src2_ibr value the result is light more than 64 Otherwise the result is dim less than 64 Grayscale subtraction subtracts the graylevel values in src2_ibr from srcl_ibr and clips the result at 0 so negative values are not created src1_ibr and src2_ibr must be in different frame stores If src1_ibr is the same as src2_ibr a binary subtraction simply zeros the binary frame store A grayscale subtraction would fill the frame store with a uniform brightness of 64 The smaller the area of the image to be processed the faster the subtraction executes Example Subtract virtual frame buffers 11 and 12 and store the result in frame 11 VSUBTRACT 1 1011 1011 1012 Related Keywords 254 VADD program instruction VCOPY program instruction VEDGE program instruction VTHRESHOLD program instruction Adept Vision Reference Guide Rev A System Switch V SUBTR
293. raphics are visible in this mode In response to VPICTURE or VWINDOW the objects in the camera view are graphically depicted in the Vision display window The image looks like display mode 2 except that only processed regions are shown Unprocessed regions those not shown are smaller than V MIN AREA larger than VMAX AREA outside the image bounds V FIRST COL V LAST COL etc In addition to the regions shown image boundaries prototype boundaries or clear grip positions are displayed depending on what switches are enabled The relevant system switches are V SHOW BOUNDS V SHOW EDGES V SHOW VERIFY V SSHOW RECOG and V SHOW GRIP Note that if too many of these display switches are enabled the resulting screen image may become muddled The VPICTURE instruction and monitor command automatically erases the vision graphics window when the display mode is 3 or overlay is in effect The erasure occurs simultaneously with the acquire and any subsequent processing Static graphics This is a graphics display mode like 3 except that the vision display window is not automatically erased with each VPICTURE Further the vision system does not automatically draw graphics when operations such as VPICTURE VRULER and VWINDOW are performed Only the graphics instructions such as GCLEAR and GLINE affect the vision display window Thus with this display mode the user has complete control over the graphics overlay image For example statistical
294. rayscale image in the destination image buffer region is left unmodified VADD type 2 adds the graylevel pixels in the two images together and divides each pixel sum by 2 to produce an average The division operation rounds up That is 1 2 is 1 2 2 is 1 3 2 is 2 4 2 is 2 etc Averaging images may be useful for filtering out noise or for increasing pixel precision VADD type 3 adds the graylevel pixels in the two images in effect increasing the image intensity Pixel sums are clipped at the maximum graylevel of 127 That is if the sum of two pixels exceeds 127 the sum is reduced to 127 Summing images may be useful for increasing the brightness of dim images Note however that the system parameters V GAIN and V OFFSET provide a more conventional method of increasing the dynamic range of images when acquiring them from the camera With VADD types 2 and 3 the binary image associated with the destination image buffer region is created by applying the threshold parameters for the given virtual camera When the two source image buffer regions can be the same the following special considerations apply e VADD type 1 is not performed This would simply OR the image with itself with no effect and copy the result to the destination image buffer region e VADD type 2 averages the image with itself again an operation with no effect but the image is copied to the destination image buffer region e VADD type 3 is the only mea
295. rayscale frame stores see the VSELECT program instruction Adept Vision Reference Guide Rev A 215 VRULERI Program Instruction Ruler types 2 1 0 2 and 3 are edge rulers and they all return the same type of information The only differences are the sources of the edge information Type 2 dynamic binary rulers use the grayscale frame store This is the data visible in VDISPLAY mode 1 The current values of the threshold parameters V THRESHOLD and V 2ND THRESH at the time the ruler is executed are used to determine which pixels are considered foreground and which are background Therefore a different binary threshold can be used for each ruler Type 1 raw binary rulers use the raw binary frame store This is the data visible in VDISPLAY mode 2 This is binary or edge data depending on the setting of the V BINARY system switch Ruler type 0 run length binary uses processed binary data produced by VWINDOW or VPICTURE in modes 1 or 0 This data is shown in VDISPLAY mode 3 Note that system parameters such as V MIN AREA and V MAX AREA are used during picture processing whereas they are not taken into account with ruler type 1 Again the image data is binary or edge data depending on the setting of the V BINARY system switch The edges returned by a type 2 fine edge ruler are based on the graylevels on and near the path of the ruler The edge detection technique used is more sophisticat
296. rder of the area of interest 1 no draw 0 erase 1 draw solid 2 complement 3 draw dashed 4 complement dashed The default is 1 draw solid Not currently used Integer value specifying the image buffer region to receive the image data that has been modified with a morphological operation Image buffer regions specify both a size and a frame store see the description of VDEF AOI Optional variable name that receives the sum of binary pixels in the destination image buffer region This can be used as a stopping criterion for edge thinning operations to indicate stasis state If count is specified forward processing is suspended until the operation completes Integer value specifying the image buffer region to apply the morphological operation The region s AOI must have been defined with a VDEF AOI instruction Binary morphological operations are nonlinear transformations of binary or edge images The VDEF MORPH instruction defines the operation to perform The VMORPH program instruction performs the operation 180 AdeptVision Reference Guide Rev A Program Instruction VMORPH Morphological operations may be used to eliminate small holes and gaps from the image by dilating and then eroding or vice versa to thin edges to isolate certain features such as straight lines etc Multiple operations are often performed in sequence Morphological operation types 1 and 2 are predefined to be erosion and d
297. re to disable the entire object recognition process when a full VPICTURE type 1 or 0 or a VWINDOW instruction is performed Connectivity will be performed to extract regions from the image but their centroids and moments will not be calculated Only the more primitive region features such as area and bounding box are computed Example Minimize vision processing for all cameras DISABLE V BOUNDARIES Related Keywords VFEATURE program instruction V FIT ARCS system switch V RECOGNITION system switch VWINDOW program instruction 38 Adept Vision Reference Guide Rev A System Switch V CENTROID Syntax V CENTROID camera Function Enable computation of the centroid of each region in the image Usage Considerations The V BOUNDARIES system switch must also be enabled in order for centroids to be computed A change to this switch takes effect when the next VPICTURE command or instruction or VWINDOW instruction is executed This is an array of switches one for each virtual camera See the general description of switches in the AdeptVision User s Guide for syntax details Details V CENTROID enables the computation of the centroids of regions When it has been computed the centroid of a region is available after a VLOCATE instruction has succeeded by using the VFEATURE function The positions and areas of holes are not considered in the centroid computation The V OSUBTRACT HOLE system
298. re the distances from the centroid to the points measured in millimeters ITEMS 48 to 50 After VLOCATE only These are the dimensions of the best fit ellipse to the region System switches V 2ND MOMENTS and V CENTROID must have been enabled for this computation Item 48 is the direction of the region major axis axis of least inertia in the range 90 to 90 degrees Items 49 and 50 are the radii of the best fit ellipse for the region The ellipse is centered at the region centroid and its major axis coincides with the region major axis See the description of the V 2ND MOMENTS switch for an explanation of the derivation Table 2 2 VFEATURE Function Data for ObjectFinder following VLOCATE Index Value Units comment 1 Valid TRUE FALSE 2 X millimeters 3 Y millimeters 4 Z millimeters 5 RX degrees 6 RY degrees 7 RZ degrees 8 Encoder offset encoder counts 9 Verify percentage percentage 11 Model number in planning list 1 10 12 Instance ID integer 13 Min_X millimeters 14 Max_X millimeters 15 Min_Y millimeters 102 AdeptVision Reference Guide Rev A Real Valued Function VFEATURE Table 2 2 VFEATURE Function Data for ObjectFinder following VLOCATE Index Value Units comment 16 Max_Y millimeters 18 Time seconds 23 Virtual camera number 1 32 27 features in model integer See the descriptions following Table
299. rea Maximum region area for recognition Effort level Effort level associated with the prototype Verify percentage Minimum percentage of the weighted prototype boundary that has corresponding image edges before the prototype is accepted as a correct match Camera associations List of the virtual cameras associated with the prototype Symmetry Special information that is displayed if the part is symmetric Subprototypes List of the subprototypes associated with the prototype AdeptVision Reference Guide Rev A 223 VSHOW Example Monitor Command Display the outline of the prototype flange in the Vision display window together with gripper definition 2 two subprototypes are associated with the prototype right_side and top part VSHOW flange 2 Color Number of taught examples Area Minimum area Maximum area Effort level Verify percentage Camera associations subprototypes White 10521 6000 14000 1 2 4 right_side top part List all the prototypes currently loaded in the system VSHOW 0 5 10 15 CASTING FLANGE ES This indicates that FLANGE and CASTING are the only object prototypes defined Their virtual camera associations are as shown Related Keywords 224 VDEESUBPROTO program instruction VEDGE INFO program instruction VGAPS program instruction VSHOW program instruction VSUBPROTO program instruction Adept Visio
300. rea to be searched for edge points Bits 4 5 Center 0 inner 8 outer 16 mask value 24 This defines the initial search point for edges The search for edges is along radial lines from the center of the arc The length of each edge search is rr If more than one edge is found along one radial line these bits 4 and 5 of the mode parameter determine which one to use By default the edge closest to the center of the search area the radius dim3 of the area of interest definition from the center dim1 and dim2 of the area of interest definition is chosen This is fine for most situations However if nearby edges are known to be present that do not belong to the arc edge one of these mode bits may be set to help avoid use of the non arc edges If inner bit 4 is set the edge closest to the center of the arc s circle is used If outer bit 5 is set the edge farthest from the center is used Only one bit may be set The VFIND ARC operator is displayed in the Vision display window as a ring or ring segment The outer half of the ring is either dark blue or light blue depending on the given color of the arc s circle mode bit 1 The inner half is drawn in the other color Bisecting the width of the ring is a guide arc drawn in green this is the estimated arc defined by elements 1 2 3 5 and 6 of the area of interest definition The inner and outer arcs show the search range third element of
301. relation using template 3 and display the search window and results The search window is centered at the point 140 260 and is 20mm x 30mm in size VDEF AOI 3000 1 140 260 20 30 VCORRELATE results 3 3011 Related Keywords VDEEFAOI program instruction VLOAD monitor command VLOAD program instruction VSHOW MODEL program instruction VSELECT program instruction VSTORE monitor command VSTORE program instruction VTRAIN MODEL program instruction AdeptVision Reference Guide Rev A 49 VDEEAOI Program Instruction Syntax VDEF AOI aoi shape diml dim2 dim3 dim4 angl ang2 Function Define an area of interest AOI Areas of interest are used by most vision tools to specify the tool placement within an image Usage Considerations A special AOI accessed as numbers 1000 1001 or 1002 is predefined for compatibility with programs written for 10 x versions of AdeptVision AGS This AOI refers to the entire frame store and cannot be redefined with VDEF AOL Parameters aoi Real value variable or expression interpreted as an integer in the range 2000 to 200000 maximum the actual range is specified with the DEVICE instruction that specifies the area of interest being defined shape Real value variable or expression specifying the shape of the area of interest See Details below diml dim4 Dimensions of the area of interest The interpretation of these values as well as which valu
302. ring and 12 bounded areas are found in the window the right most 2 areas are ignored That is VOCR reports that only 10 characters were found If fewer character regions are found than expected the expected characters with no corresponding bounded areas are given a score of 0 when computing the average and minimum scores Also data i 3 expected characters were all 1st picks and data i 4 they were all 1st or 2nd picks are both assigned the value FALSE Merged characters are automatically split when there is expected text This includes the VOCR operations Quick verification op 0 and Robust verification op 1 or 2 but not Recognition op 3 Splitting works with rotated VOCR windows The splitting criteria are simple First there must be fewer regions in the VOCR window than expected Then while processing the bounded areas in the VOCR window from left to right the following test is performed If the width of the current area is closer to the total width of the current and next expected characters than to the width of the character currently expected then the bounded area is split AdeptVision Reference Guide Rev A Program Instruction VOCR A bounded area is split based on the average width of the character currently expected The average width is based on the trained instances of the character This simple strategy for splitting works well when only two characters are merged int
303. riptions are the same except for the flags bit returned by VFEATURE that indicates if the hole boundary contributed to the recognition of the object VLOCATE in find hole mode refers to the object most recently VLOCATEd regardless of which program task executed the VLOCATE instruction Consequently there is possible confusion when more than one program task performs VLOCATEsS Thus for predictable operation with find hole requests applications should be organized to have only one program task execute VLOCATE instructions The other VLOCATE mode bits are ignored in find hole mode That is no wait mode and find any mode are implicitly in effect Also the name parameter to VLOCATE must be omitted when find hole mode is specified VLOCATE in find hole mode succeeds only after a VPICTURE or VWINDOW processed with the V HOLES switch enabled and the V DISJOINT switch disabled Default Find object mode 0 The order parameter to VLOCATE allows you to locate objects in the image in a preferred order biggest first left most first etc This applies to recognized objects unrecognized regions and holes depending on the mode of the VLOCATE The order parameter must have a value in the range 0 to 10 The values are interpreted as follows Value Object Indicated Any Biggest first not used for ObjectFinder Smallest first not used for ObjectFinder Left most based on box center Right most based on box c
304. rithm The effort level specifies how much time should be spent searching for edge points If the effort level is 1 only a few edge points about five evenly spaced across the angular range are sought If the effort level is 100 as many edge points as possible are sought High effort levels with big arcs use a lot of execution time Unless the arc image is small or noisy a low effort level should be used This should provide adequate accuracy VFIND ARC optionally filters out edge points that are far from the fit arc and then refits the arc This is an iterative process that increases the execution time but provides more accurate and consistent results particularly when noise is present in the image Two system parameters V MAX SD and V MAX PIXEL VAR control the filtering process VFIND ARC filters edge points if VMAX SD is nonzero AdeptVision Reference Guide Rev A 111 VFIND ARC Program Instruction 112 Otherwise filtering is disabled V MAX SD specifies the number of distance standard deviations from the fit arc beyond which edge points should be discarded V MAX PIXEL VAR specifies the maximum pixel distance below which no edge points should be discarded See the description of the V MAX SD system parameter for more information on the iterative filtering process The edge points found by VFIND ARC are displayed in the Vision display window when the V ISHOW EDGES system switch is enabled Edge points used in the final fit of the
305. rogram instruction VCORRELATE program instruction VDEEFONT program instruction VDELETE monitor command and program instruction VOCR program instruction VSHOW MODEL program instruction Adept Vision Reference Guide Rev A 281 VWAIT Program Instruction Syntax VWAIT type ibr Function Delay program execution until processing of a VPICTURE or VWINDOW opera tion is complete Usage Considerations VWAIT is not considered during vision model training Parameters type Optional integer expression having one of the following values 0 Wait for the vision system to become idle default 1 Wait for the acquire into the frame specified in ibr to complete 2 Wait for the acquire into the frame specified in ibr to start ibr Optional integer value specifying an image buffer region Only the frame store element of the image buffer region is used see the description of VDEF AOI This parameter is ignored for type 0 VWATTs Details A type 0 VWAIT waits for the vision processor to become idle following a VPICTURE or VWWINDOW operation A type 0 VWAIT is necessary before executing any of the following instructions VDEF FONT VDEF MORPH VGETPIC VHISTOGRAM VPUTCAL VPUTPIC VSHOW VSHOW MODEL VSUBPROTO VTRAIN MODEL VDELETE monitor command only A type 2 VWAIT is essentially a wait for a strobe light or camera shutter to operate 282 AdeptVision Reference Guide Rev A Program Instruction VWAIT Example V
306. rom disk windows but not with rotated windows V BORDER DIST is CAUTION V BORDER DIST works with orthogonal VWINDOW ignored if the window is rotated Example Figure 2 1 shows an image of a pallet containing round objects Many of the objects have been clipped by the image border To speed up image processing the V BORDER DIST parameter was assigned the value 25 This defines an inner border that is depicted by dashed lines in the figure Features outside the inner border are not used to propose or confirm matches The features may still be used to verify matches however 36 AdeptVision Reference Guide Rev A System Parameter V BORDER DIST gt V BORDER DIST 25 Figure 2 1 Example of VBORDER DIST Related Keywords VDEF AOI program instruction V FIRST COL system parameter V FIRST LINE system parameter V LAST COL system parameter V LAST LINE system parameter AdeptVision Reference Guide Rev A 37 V BOUNDARIES System Switch Syntax V BOUNDARIES camera Function Enable or disable boundary analysis by the vision system Usage Considerations A change to this switch takes effect when the next VPICTURE command or instruction or VWINDOW instruction is executed This is an array of switches one for each virtual camera See the general description of switches in AdeptVision User s Guide for syntax details Details This switch allows the user to disable boundary analysis and therefo
307. rom the source area of interest it has already been written over by line 1 of the source area of interest Examples Copy one image buffer region in virtual frame store 11 to another image buffer region in the same virtual frame store VCOPY 4011 5011 Related Keywords 44 VDEFAOI program instruction VDEELUT program instruction VADD program instruction VSELECT program instruction VSUBTRACT program instruction AdeptVision Reference Guide Rev A Program Instruction VCORRELATE Syntax VCORRELATE cam mode dmode max_depth accept give_up data i act_depth tplnum ibr VCORRELATE cam mode dmode data i tplnum shape cx cy dx dy ang Function Perform a normalized grayscale or binary correlation comparing a predefined template with a rectangular image window or searching for a closest match within the window to a predefined template Usage Considerations Binary correlation requires the AdeptVision Enhanced VXL Interface option Templates are defined with a VTRAIN MODEL instruction Parameters cam Optional real valued expression indicating the virtual camera number to use This parameter is currently not used mode Optional real valued expression used to specify a skip pattern as follows 0 default use the mode determined by the vision system 1 nonskipping mode 2 skip rows only 3 skip columns only 4 skip both rows and columns max_depth Optional real valued expression indicating
308. rototype is referenced AdeptVision Reference Guide Rev A Program Instruction VSUBPROTO If subproto_name is not specified edge_num must be specified and cannot be zero and refers to an edge in the last prototype VLOCATEd using the edge numbering for that prototype Details After a prototype has been recognized and VLOCATEd the VSUBPROTO instruction may be used to find what percentage of a subprototype or individual edge was verified The verify percentage returned is an unweighted quantity That is it is not adjusted by the edge weights that the user is able to assign during training This instruction also can be used to refine the position information for the prototype based on only a subprototype or a single edge The adjusted location for the prototype is returned via the parameter trans_var The adjustment computation is based on the final verification of the prototype Information is retained in the vision system recording which portion of each prototype edge was verified by which portion of each image edge The position refinement calculations account for the lengths of the edges verified their positional variance based on the training instances the verified corners of the edges and their positional variances If you are interested in only one edge or a few edges the VFIND LINE and VFIND ARC instructions provide greater subpixel location accuracy You can use the VSHOW command or instruction to d
309. s A large template with a size of 400x400 pixels will use 160 000 bytes of vision CPU memory The FREE monitor command and program instruction report both the amount of vision CPU memory available and the amount used for all models templates fonts and prototypes OCR Fonts After a font has been defined with the VDEF FONT instruction the characters in the font can be trained The vision system is taught to recognize characters by showing it what they look like Each character should be shown to the system a number of times because the system accumulates statistics on the appearance of the characters See the AdeptVision User s Guide for an overview of the OCR capability in AdeptVision Adept Vision Reference Guide Rev A 277 VTRAIN MODEL Program Instruction 278 To train the vision system on characters you must define an image buffer region that encompasses them in the current image Together with the image buffer region you tell the vision system what characters are inside the region ordered left to right as you would normally read the text Since the region can be rotated the left edge of the window is drawn a different color for reference For a rotation of 180 degrees the left edge is actually on the right and the text inside should look upside down When the vision system analyzes the bounded areas in the image buffer region it orders the bounded areas by distance from the left edge The characters specifi
310. s VFEATURE 96 VQUEUE 210 Related publications 8 S Sample Line Finder tool 120 Service calls 14 Shape parameters for arc shaped tools 53 for rectangular tools 51 Sobel edge operator 94 Support application support 15 Internet E Mail Address 15 phone numbers 14 training information 15 System parameters V2ND THRESH 22 V BORDER DIST 35 V EDGE STRENGTH 92 V EDGE TYPE 94 V FIRST COL 129 V FIRST LINE 130 Adept Vision Reference Guide Rev A V GAIN 132 VIO WAIT 149 V LAST COL 152 V LAST LINE 153 V LAST VER DIST 154 V MAX AREA 165 V IMAX PIXEL VAR 167 V IMAX SD 170 V MAX TIME 172 V MAX VER DIST 173 V MIN AREA 175 V MIN HOLE AREA 176 V MIN LEN 177 V MIN MAX RADI 178 V OFFSET 189 VSYNCH STROBE 257 V THRESHOLD 262 System switches T V 2ND MOMENTS 20 V BACKLIGHT 31 V BINARY 33 V BOUNDARIES 38 V CENTROID 39 V DISJOINT 77 V DRY RUN 84 V FIT ARCS 131 V HOLES 148 V OVERLAPPING 191 V PERIMETER 192 V RECOGNITION 211 VSHOW BOUNDS 229 VSHOW EDGES 230 V SHOW FEATS 232 V SHOW GRIP 233 VSHOW RECOG 237 V SHOW VERIFY 238 V STROBE 248 V SUBTRACT HOLE 255 V TOUCHING 263 VISION 151 Threshold setting second binary 22 Training information 15 U Upgrading program code 313 Index V V Language Reference Guide 9 V Language User s Guide 9 V 2ND MOMENTS system switch 20 V 2ND THRESH system parameter 22 V BACKLIGHT system switch 31 V BINARY system switch 33 V BORDER DIST system parameter 35 V BOUNDARIES sy
311. s and corners of a prototype or of a region in the image V EDGE INFO camera Enable saving of information about edges in the image for recall via the VEDGE INFO instruction V EDGE STRENGTH camera Set the edge threshold for grayscale image processing and fine edge rulers V EDGE TYPE camera Determine the type of edge operator to use cross gradient or Sobel when a VPICTURE instruction is performed VFEATURE index Return specified information about the object most recently VLOCATEd or the prototype most recently displayed by the VSHOW program instruction VFIND ARC cam mode dmode effort type data i ibr VFIND ARC cam mode dmode effort type data i 1 xc yc r rr ang0 angn Fit a circular arc to an image edge bounded by a window that is shaped like a ring or a ring segment VFINDER cam type dmode how_many_total times ibr VFIND LINE cam pos dmode effort type data i ibr VFIND LINE cam pos dmode effort type data i 1 xc yc length width angle Fit a straight line to an image edge within a window VFIND POINT cam pos dmode effort type data i ibr VFIND POINT cam pos dmode effort type data i 1 xc yc length width angle In a search window find the edge point that is nearest to one side of the window V FIRST COL camera Set the number of the first column of pixels to be processed V FIRST LINE camera Set the number of the first li
312. s are not square or there is some tilt to the camera The vision system compensates for this so that the window covers a square area in the scene For shape 2 windows the width and height are the X radius and Y radius respectively of the outer edge in pixels This reflects the millimeter per pixel ratio and the X Y ratio in the calibration Type 0 windows differ from all the other types in that they are applied to the frame grabbed binary image that you see in VDISPLAY mode 2 Type 0 windows simply return a count of the white pixels in the window If the V BINARY system switch is enabled this is the number of background or foreground pixels depending on the physical lighting setup If V BINARY is disabled this is the number of edge pixels in the window All windows except type 0 are applied to the frame grabbed grayscale image you see in VDISPLAY mode 1 Window types 1 2 3 and 4 return the average graylevel for the pixels in the window Types 2 3 and 4 also return the minimum and maximum graylevels found in the window For types 3 and 4 you also get a count of the background and foreground pixels as determined by the V THRESHOLD and V 2ND THRESH system parameters and the V BACKLIGHT system switch for the virtual camera specified The criterion for classifying a pixel as being background or foreground is the same as that used during a VPICTURE operation see the descriptions of V THRESHOLD and V BACKLIGHT For ty
313. s specified its prototype image is displayed in the Vision display window and a list of its descriptor values is displayed in the Monitor display window along with the names of any subprototypes associated with the prototype The prototype is displayed at its nominal location in the Vision display window This is the location of the first instance of the prototype If display of a subprototype is requested the prototype is displayed in gray and the subprototype in color If the grip parameter is nonzero the specified grip position is shown overlaying the prototype in the Vision display window If edge_num is nonzero one or more edge numbers are shown You may need to know these edge numbers for the program instructions VEDGE INFO VDEFSUBPROTO VSUBPROTO and VGAPS The following are explanations of the descriptor values displayed Color Color of the object when it was trained During subsequent image processing in order for a region to be recognized as an object it must have the same color as the prototype The value of color will be either Black or White Number of taught Number of images that were used during training to examples generate the prototype The more images used the more reliable the prototype Area Area of the prototype in camera pixels including any holes This is the area of the first trained instance of the prototype Minimum area Minimum region area for recognition Maximum a
314. sing and all pending vision operations associated with the given task number VADD Add two binary or grayscale images VAUTOTHR Determine good thresholds for binary image processing based on the gradients in a grayscale frame store V BACKLIGHT Establish the color black or white of the background V BINARY Enable or disable automatic edge image generation at VPICTURE time V BORDER DIST Define an image border reduction in pixels to mask out regions clipped by the image border V BOUNDARIES Enable or disable boundary analysis by the vision system V CENTROID Enable computation of the centroid of each region in the image VCONVOLVE Perform an image convolution on a grayscale frame possibly storing the result in a different frame store VCOPY Copy all or part of an image from one operation region definition to another Adept Vision Reference Guide Rev A 9 Chapter 1 AdeptVision Keyword Summary Table 1 2 AdeptVision Keyword Summary Continued Keyword Description VCORRELATE Perform a normalized correlation comparing a predefined template with a rectangular image window or searching for a closest match within the window to a predefined template VDEF AOI Define an area of interest Areas of interest are used by most vision tools to specify the tool placement within an image VDEF CONVOLVE Define an image convolution VDEF FONT Define replace
315. spective distortion the correction is applied to the placement of the ruler Since perspective transformations of lines are lines the two end points of a linear ruler are transformed to the perspectively corrected plane Thus the results from the ruler millimeter distances to edge points along the ruler are automatically corrected for distortion The center of an arc ruler is similarly transformed However the path of an arc ruler that is followed in the image memory is an upright ellipse This only approximates the perspectively distorted arc A perspectively distorted circle is neither a circle nor an ellipse The two radii of the ellipse are determined by the X Y ratio in the camera calibration data The angles to edges will be correct in the perspectively corrected plane but edge points computed based on those angles using the center and radius of the arc ruler will not be exact because the ruler shape is not perfectly elliptical when there is perspective distortion Examples 218 Cast a run length type 0 binary ruler along a horizontal line starting at the point 30 20 of length 80 and put the detected binary transition information into the array xx For maximum speed don t draw the ruler dmode 1 and stop after finding two edges maxcnt 2 VRULERI 0 1 2 xx 1 30 20 80 AdeptVision Reference Guide Rev A Program Instruction VRULERI Cast a fine edge type 2 ruler along the same line as
316. ssion indicating the type of data to store in the array pic The default is 1 1 Grayscale image and binary or edge image 2 Binary or edge image only s_rate Optional real valued expression specifying the sampling rate for reading pixels If the rate is N every Nth pixel is read out of every Nth row The rate must be in the range 1 to 100 The default is 1 s_mode Optional real valued expression specifying the sample mode which is meaningful only if the sample rate s_rate is greater than 1 Pixels are simply sampled if s_mode is 0 the default Pixels are averaged if s_mode is 1 in which case the binary portion of the image will be stripped NOTE The parentheses in the instruction syntax can be omitted if all four of the above parameters are omitted pic Array into which to put a header string and picture data rane Row and column indexes into the array pic indicating where the picture and header information is to be stored The defaults are 0 for both r and c The header string is stored in the element pic r c The picture data is stored starting at element pic r 1 c 1 shape Optional real valued expression indicating the shape of the image to read The default is 1 indicating a rectangular shape which is the only shape now available Adept Vision Reference Guide Rev A 141 VGETPIC Program Instruction x0 y0 Optional real valued expressions specify
317. ssion specifying the type of line finder 2 dynamic binary 1 raw binary 0 run length binary 2 fine edge The default is 2 fine edge Real array describing the outcome of the line fit data i 0 TRUE if a line was fit otherwise FALSE data i 1 TRUE if any part of the search window falls off the vision display window data i 2 data i 3 X Y coordinate on the line nearest to the initial search point in millimeters AdeptVision Reference Guide Rev A ibr xc yc length width angle Details Program Instruction VFIND LINE data i 4 Angle of the fit line in degrees data i 5 Percentage of the guidelines extent for which edge points were found 0 0 to 100 0 data i 6 Maximum error distance from the fit line to the most distant edge point found in pixels data i 7 Maximum error toward the dark side of the line in pixels data i 8 Maximum error toward the bright side of the line in pixels data i 9 Percent of edge points filtered out Integer value specifying the image buffer region within which to search for a line Image buffer regions specify both a size and a virtual frame buffer see the description of VDEF AOI Optional array index that identifies the first element to be defined in data The default is 0 If a multiple dimension array is specified only the right most index is incremented as the values are assigned Real valued expressions specifying the cente
318. stem switch 38 V CENTROID system switch 39 V DISJOINT system switch 77 V DRY RUN system switch 84 V EDGE INFO program instruction 90 V EDGE STRENGTH system parameter 92 V EDGE TYPE system parameter 94 V FIRST COL system parameter 129 V FIRST LINE system parameter 130 V FIT ARCS system switch 131 V GAIN system parameter 132 V HOLES system switch 148 V IO WAIT system parameter 149 V LAST COL system parameter 152 V LAST LINE system parameter 153 V LAST VER DIST system parameter 154 V MAX AREA system parameter 165 V MAX PIXEL VAR system parameter 167 V MAX SD system parameter 170 V MAX TIME system parameter 172 V MAX VER DIST system parameter 173 V MIN AREA system parameter 175 V MIN HOLE AREA system parameter 176 V MIN LEN system parameter 177 V MIN MAX RADII system parameter 178 V OFFSET system parameter 189 V OVERLAPPING system switch 191 V PERIMETER system switch 192 V RECOGNITION system switch 211 V SHOW BOUNDS system switch 229 V SHOW EDGES system switch 230 V SHOW FEATS system switch 232 V SHOW GRIP system switch 233 V SHOW RECOG system switch 237 V SHOW VERIFY system switch 238 V STROBE system switch 248 V SUBTRACT HOLE system switch 255 Adept Vision Reference Guide Rev A 321 Index V SYNC STROBE system parameter 257 V THRESHOLD system parameter 262 V TOUCHING system switch 263 VABORT monitor command and program instruction 24 VADD program instruction 25 VAUTOTHR monitor command and program instruction 28 VCONVO
319. system is loaded The maximum physical camera number is 4 instead of 8 5x5 and larger convolutions can not be applied on full frame 640x480 image They are limited to 508 pixel wide AOIs Correlation templates and the AOIs of image processing operations are reduced if necessary to become a multiple of 4 pixels in width If you need to know the exact size of a template use VSIHOW MODEL V SHOW MODEL now shows templates in the middle of the vision window AdeptVision Reference Guide Rev A Appendix D Compatibility Summary VDEF CONVOLVE must be given a 7x7 array instead of a 5x5 array VPUTPIC strips the binary out of averaged images Type 0 stores both grayscale and binary images Instructions with added arguments that make them incompatible are VADD VCONVOLVE VEDGE VMORPH VSUBTRACT VTHRESHOLD These incompatibilities are caused by standardizing on the cam type dmode argument triples VFIND LINE VFIND ARC and VFIND POINT have a new shape argument for consistency with the other instructions Full frame and field only image acquires do not mix well If you plan to do both with the same vision system you experience some delay when switching back and forth The reason for this is that full frame is interlaced and field only is noninterlaced yet all cameras connected to the MUX are being driven by the same synch signals VWINDOW VOCR VTRAIN MODEL on a font and VTRAIN automatically do VSELECT of the sp
320. t from the region centroids This information is available with the VFEATURE function after a VLOCATE instruction has succeeded The points are represented by their distances and directions angles from the region centroid This information is also available for holes To get information about holes you should enable V HOLES disable V DISJOINT and do VLOCATEs in get hole mode Example For each region in the image the following program segment draws lines from the centroid to the points on the perimeter that are closest to and farthest from the centroid Required switches ENABLE V BOUNDARIES V CENTROID V MIN MAX RADII VDISPLAY 3 Special display VPICTURE 0 Take a picture with virtual camera 1 and no recognition ATTACH vlun 4 GRAPHICS Attach to the vision window FOPEN vlun Vision MAXSIZE 640 480 amp select graphics scaling GTRANS vlun 1 in real world millimeters vfi cx 42 Indexes of VFEATURE function v cy 43 for centroid vf minr ang 44 and min max radii values 178 Adept Vision Reference Guide Rev A System Switch V MIN MAX RADII vf maxr ang 45 vf minr dist 46 vf maxr dist 47 VWAIT Wait for image processing to complete for graphics instr VLOCATE Snam Locate anything in the image WHILE VFEATURE 1 DO If
321. tVision Reference Guide Rev A 159 VLOCATE Syntax Program Instruction VLOCATE camera mode order name trans_var Function Identify and locate an object in the scene Parameter camera mode order Optional real valued expression that specifies the virtual camera number All cameras are implied if the camera number is 0 Optional bit field expression indicating whether a particular object is being sought and whether the VLOCATE should wait if necessary See below for details Optional real valued expression for selecting objects by size or position in the image The default is 0 for no specific order desired See below for details NOTE This parameter has certain restrictions for the ObjectFinder tool See the table on page 162 for details Sname trans_var Details String variable to be assigned the name of the object found find any mode or a string expression representing the name of the object to be located find particular mode This parameter must be omitted in find hole mode see below Optional transformation variable to be assigned the location of the object See details below The vision system keeps a queue of the recognized objects and unrecognized regions found during image processing The queue is filled by doing VPICTURE or VWINDOW instructions and is emptied using the VLOCATE instruction Each VPICTURE and VWINDOW operation clears the vision queue for the specifie
322. ted the model information is written in a special format for later recall with the VLOAD command or instruction Examples The following command stores all the prototypes from the vision system into a file named OBJECTS VS This file will contain two prototypes casting and flange with the virtual camera associations shown VSTORE objects 0 5 10 15 20 25 30 ssl ls lr lo ls ls CASTING FLANGE The following command stores the prototypes smd1 and quad ina file named PROTOS VS on disk drive B VSTORE B protos smdl quad 0 5 10 15 20 25 30 dae A ls ee heras hea ll eae os SMD1 Kk QUAD k Related Keywords VLOAD monitor command VLOAD program instruction VSTORE program instruction 244 AdeptVision Reference Guide Rev A Syntax Program Instruction VSTORE VSTORE lun file_spec model_name Smodel_name Function Store in a disk file selected or all vision prototypes and their subprototypes Optical Character Recognition OCR fonts or correlation templates Parameters lun file_spec Smodel_name Real valued expression that specifies the logical unit number to be associated with the operation This must be one of the logical unit numbers for a disk device see the ATTACH instruction in the V Language Reference Guide The logical unit number used must not already be in use by the program for another disk access String expression
323. ter file _spec Specification of the disk file from which the vision models are to be loaded This consists of an optional physical device an optional disk unit an optional directory path a file name and an optional file extension Uppercase or lowercase letters can be used The current default device unit and directory path are considered as appropriate see the DEFAULT command in the V Operating System Reference Guide If no filename extension is specified the extension VS is appended to the name given Details All the vision models previously VSTOREd in the given file are loaded and added to those already in memory If the file contains a vision model with the same name font number or template number as one already in the vision system loading is aborted That is none of the vision models in the file are loaded Any subprototypes associated with a prototype are also restored automatically As the VLOAD command is processed V displays in the Monitor display window the names of the models along with their virtual camera associations If fonts or templates are loaded their identifying numbers are displayed 156 AdeptVision Reference Guide Rev A Monitor Command VLOAD Example The following monitor command loads the models from the disk file named OBJECTS VS and displays their virtual camera associations as shown VLOAD objects 0 5 10 15 20 25 30 less less ls ler lr laa les CASTING
324. ter models if the number of regions in the window is not identical to the number of characters in the given text parameter However if you make a mistake that does not result in VTRAIN MODEL returning an error you should VDELETE the font and start over by redefining the font Each character must appear as a single bounded binary area Thus for example the characters and cannot be trained The V MIN AREA system parameter should be adjusted to filter out the dots over the letters i and j Also if characters touch the system processes them as a single character Automatic splitting of touching characters is not done during training VTRAIN MODEL displays the outline of each bounded area in the Vision display window so you can see the bounded areas that were analyzed Before training on a font you should run the program set gain in the file SET_GAIN V2 on the Adept Utility Disk to adjust the gain and offset Next take a picture and find a good threshold AdeptVision Reference Guide Rev A Program Instruction VTRAIN MODEL using the VAUTOTHR monitor command or with the pull down menus You should also take some sample pictures in VDISPLAY mode 3 and adjust the parameter V MIN AREA to filter out dots and noise Keep V MIN HOLE AREA low enough so that the holes in letters do not disappear Characters may be trained in one orientation and recognized in a different one In fact characters in a font ma
325. ters one for each virtual camera See the general description of parameters in the AdeptVision User s Guide for syntax details Details If the V IO WAIT parameter is setto 1 VPICTURE commands or instructions wait for an interrupt from the fast digital input interrupt line before acquiring an image This is useful for taking pictures of fast moving objects For example the object could trigger a simple sensor such as a photoelectric cell that is wired into the V controller NOTE When taking pictures of moving objects a strobe light or shuttered camera should be used to minimize image blur For more information on the use of strobe lights see the system switch V STROBE and the system parameter V SYNC STROBE in this manual For information on the various camera types see the AdeptVision User s Guide The fast digital input interrupt line provides a significant improvement in response time to external events When this line is used the worst case delay from the time the line is triggered until the strobe light output signal is fired is about 10 microseconds If instead a V program uses the WAIT instruction to wait for a digital signal before performing a VPICTURE the delay can be as long as 16 milliseconds The worst case delay should be planned for if objects in the scene are moving fast with respect to the camera field of view This worst case delay assumes the Adept system controls a robot has multiple execution tasks
326. th split images where one half of the image is calibrated differently from the other half If a quick frame grab is still acquiring an image into the frame store being selected the VSELECT instruction causes program execution to wait for the acquisition to complete For some vision operations such as binary correlation two scratch frame stores are needed The display frame store will always be used for one For the other the system will use any frame store that does not contain valid image data However if all frame stores contain valid data then an error code will be returned Once a frame store has valid data it will continue to have valid data until an ENABLE VISION instruction is processed VSELECT with mode 1 will deselect a frame buffer making it available as a scratch frame store the optional camera argument is ignored Related Keyword VPICTURE monitor command and program instruction VDEF AOI program instruction AdeptVision Reference Guide Rev A 221 VSHOW Syntax Monitor Command VSHOW proto_name grip edge_num Function List the defined prototypes or display a vision prototype a subprototype or a specific prototype edge in the Vision display window Edge numbers are option ally shown Usage Considerations The VISION switch must be enabled the vision processor must be idle and vision model training must not be active for this instruction to be executed This command supports the
327. the area of interest definition for the search The initial search point is indicated by a knot in yellow on either the inner center or outer arc depending on mode bit 4 and 5 The resulting fit arc is drawn in red if one is found AdeptVision Reference Guide Rev A 109 VFIND ARC Program Instruction When using the first syntax the shape defined by VDEF AOI should be 5 6 7 or 8 Figure 2 6 shows the mechanics of an arc finder These mechanics are determined by the mode bits in the VFIND ARC instruction and the shape specification in the area of interest definition The appearance of the tool shown in Figure 2 6 was created with a VDEF AOI instruction specifying shape 5 The tool dimensions from the VDEF AOI instruction are as follows dim1 and dim2 are the center of the finder tool within the image dim3 is the guide radius shown as a dotted line in Figure 2 3 dim4 is the search range and ang1 and ang2 indicate the angular range of the search The mode bits from the VFIND ARC instruction are bit 1 is 0 indicating a search for an arc with a dark inside shown by the dark blue inner radius of the tool bit 2 0 bit 3 1 indicating that the arc s radius and center will be calculated and bit 4 0 bit 5 1 indicating that the search will start from the outer radius shown by a white dot on the outer radius The following code will create the arc finder in Figure 2 6 arc_ibr 2011 area of interest 2
328. the clear grip tests ENABLE V SHOW GRI FU Related Keyword VDEFGRIP program instruction AdeptVision Reference Guide Rev A 233 VSHOW MODEL Program Instruction Syntax VSHOW MODEL mode chars data i model_num Function Display a model either a correlation template or an Optical Character Recogni tion OCR font and return information about it or return information about all the defined templates or OCR fonts Usage Considerations The VISION switch must be enabled the vision processor must be idle and vision model training must not be active for this instruction to be executed Parameters mode Optional expression that is interpreted as a bit field to control the behavior of this instruction Currently only bit 1 mask value 1 is used If mode is 0 the font s or template s is are displayed If mode is 1 nothing is displayed The default is 0 chars Optional string variable to be assigned the set of characters in the font specified by the parameter mode1_num This parameter is ignored if model_num refers to all fonts or to a correlation template data Optional array into which font or template information is to be placed The amount of information depends on whether the model specified by mode1_num see below is a font or a correlation template The information returned further depends on whether all models that is all fonts or all templates or a single m
329. the maximum depth of a hierarchical search as follows 1 default go to the maximum depth possible see the description Of act_depth If the specified template is a binary template hierarchical searches are not allowed and this parameter is ignored O indicates no hierarchical search 1 4 depth of hierarchical search accept This parameter specifies the minimum acceptable score for the final match Default value is 0 8 Searching stops when this score is returned If a match exceeding this score is not found the best match is returned give_up If a hierarchical search is performed this parameter specifies the minimum score that must be realized at each level of the search in order to continue searching at the next higher level If a search does Adept Vision Reference Guide Rev A 45 VCORRELATE 46 dmode data act_depth tplnum ibr shape cx cy ang Program Instruction not find a match that exceeds this score the search returns to the next candidate at the lowest level Default value is 0 2 Optional real valued expression specifying the display mode for this operator The choices are 1 no draw 0 erase 1 draw solid 2 complement 3 draw dashed and 4 complement dashed The default is 1 draw solid When drawn both the search window and the highest scoring template position are shown Real array into which the results of the correlation are placed Element datali receives the
330. the mode parameter 212 Descriptions of Vision Keywords 17 Disjoint image regions special processing 77 Display frame store 54 E Edge operators cross gradient 94 Sobel 94 Edge strength setting 92 Effect of V DISJOINT switch 78 Effects of VMAX PIXEL VAR parameter 168 Elements of VGETCAL VPUTCAL scaler calibration array 204 E mail address 15 Example of VBORDER DIST 37 F Fax On Demand Adept 16 First column processed in image 129 First line processed in image 130 Fitting lines and arcs to edges 303 Fitting primitive edges to chains 303 Foreground differentiating from background 31 Frame store display 54 France Adept office 15 G Gain setting video 132 Grayscale average 25 averaging 253 camera output 79 80 correlation normalized 45 data 142 215 edges 216 frame buffers virtual 220 frame store 9 11 28 56 108 118 125 126 146 194 216 frame grabbed 79 80 293 image processing edge threshold 10 92 images 10 13 26 85 92 94 141 181 197 206 207 241 253 260 293 294 and binary 44 315 edge strength 92 mode 108 167 170 255 precision calculations 219 processing 31 262 range 22 subtraction 253 254 summed 25 tool 117 underlying image 26 253 values range of 22 H High speed trigger waiting for 149 How to use the arrays pmm to mm 318 Adept Vision Reference Guide Rev A and mm to pmm 311 I Image background vs foreground 31 Image boundaries setting bottom
331. the range 1 at the left edge of the virtual frame buffer to the current value of the parameter V LAST COL The parameter V FIRST COL is set to 1 when the V and AdeptVision systems are loaded into memory from disk Example Process all camera images starting at the left most border PARAMETER V FIRST COL 1 Related Keywords VDEEAOI program instruction V FIRST LINE system parameter V LAST COL system parameter V LAST LINE system parameter AdeptVision Reference Guide Rev A 129 V FIRST LINE System Parameter Syntax V FIRST LINE camera Function Set the number of the first line of pixels to be processed Usage Considerations A change to this parameter takes effect when the next VPICTURE or VWWINDOW instruction is executed This is an array of parameters one for each virtual camera See the general description of parameters in the AdeptVision User s Guide for syntax details Details This parameter together with V LAST LINE is used to set the range of camera pixel lines that are processed during VPICTURE or VWWINDOW operations Lines outside the range specified are ignored by the vision system VDEF AOI is the preferred method for defining a processing border for VPICTURE and VWINDOW This parameter must be assigned an integer value in the range 1 for the line at the bottom of the virtual frame buffer to the current value of the parameter V LAST LINE The parameter V FIRST LINE is set to 1 when the V
332. the verify percentage provided during prototype training Adept Vision Reference Guide Rev A 305 Perspective Distortion Overview of Perspective Distorion Calibration AMayS o How to Use the Arays pmm to mm and mm to pmm AdeptVision Reference Guide Rev A 307 Appendix C Overview of Perspective Distortion Overview of Perspective Distortion The camera calibration programs supplied by Adept relate the pixels picture units of the vision system to millimeters in your workspace These programs produce two types of calibration data scale transformations and perspective transformations The scale transformation is simple and fast The perspective transformation is more complex and a little slower but more accurate The vision system can use both the scale and perspective transformations or it can use only the scale transformation The scale transformation consists primarily of two scalers for converting pixels to millimeters Since the height and width of a pixel are not necessarily the same the X and Y components of coordinates are scaled differently The two scalers are X scale and Y scale in terms of millimeters per pixel If X Y is a coordinate in pixel units the corresponding coordinate in millimeters is X X scale Y Y scale The values for X scale and Y scale are stored in the array specified as a parameter for the VPUTCAL progr
333. then try each of the predefined convolutions by repeatedly doing VPIC 2 followed by VCON n with n equal to 1 2 16 Convolutions types 1 to 16 are predefined and types 17 to 32 can be defined by the user See the VDEF CONVOLVE instruction in this manual for a description of the predefined convolutions and an explanation of how to define new ones The image buffer region defines the convolved area The smaller the area of the image to be processed the faster VCONVOLVE executes Tf the destination buffer is different from the source buffer the destination buffer assumes some attributes of the source buffer In particular the virtual camera that was associated with the source buffer when the image was acquired via VPICTURE becomes associated with the destination buffer This affects subsequent measurement or inspection operations on the frame such as VRULERI or VWINDOW because conversion of pixel units to millimeters is determined by the camera calibration Example Apply a low pass averaging filter to the currently selected full frame buffer VCONVOLVE 1 Related Keywords VDEEAOI program instruction VDEECONVOLVE program instruction 42 AdeptVision Reference Guide Rev A Program Instruction VCOPY Syntax VCOPY cam scale dmode lut dest_ibr src_ibr Function Copy the image from one image buffer region to another Usage Considerations The scaling and look up table features require the AdeptVis
334. ting of VOVERLAPPING affects the recognition strategy used by the vision system and the way recognition proposals are verified in the image If V OVERLAPPING is enabled any edge in the image may be used to verify multiple objects not just one object For example if the corresponding holes of two overlapping objects line up the one hole seen could be used to verify both objects Since edges may be used to verify the presence of more than one object in some situations the vision system may incorrectly recognize two or more objects where there is only one A V program could be written that compares the locations of the objects in order to discard the invalid duplicates Example Do not allow objects to overlap in the image DISABLE V OVERLAPPING Related Keyword V BOUNDARIES system switch V TOUCHING system switch AdeptVision Reference Guide Rev A 191 V PERIMETER System Switch Syntax V PERIMETER camera Function Enable computation of the lengths of region perimeters Usage Considerations The V IBOUNDARIES system switch must be enabled to compute perimeters A change to this switch takes effect when the next VPICTURE command or instruction or VWINDOW instruction is executed This is an array of switches one for each virtual camera See the AdeptVision User s Guide for details on setting switches Details When this switch is enabled the outer boundary of each region in the image is scanned a
335. ting the shape of the window Currently the only choice is 1 for rectangular NOTE If any of the following four parameters cx cy width or height is specified all four parameters must be specified cx Cy width height ang ibr Optional real valued expressions specifying the center coordinate of the window in millimeters Optional real valued expression specifying the width of the window in millimeters Optional real valued expression specifying the height of the window in millimeters Optional real valued expression specifying the orientation of the window in degrees The default is 0 degrees Optional integer value specifying the image buffer region for training a prototype Image buffer regions specify both a size and a AdeptVision Reference Guide Rev A 265 VTRAIN Monitor Command frame store see the description of VDEF AOI The image buffer regions AOI must specify a rectangular shape Details VTRAIN is used to create a new object prototype or to modify an existing prototype If the prototype exists the user may show the vision system a new instance or change the prototype s virtual camera associations effort level edge weights verification threshold minimum maximum areas or position constraints A window may be specified when creating a new object prototype or training another instance of a prototype The window can be used to limit the portion of the image that is considered The wind
336. to be considered the background Usage Considerations A change to this switch takes effect when the next VPICTURE command or instruction or VWINDOW VWINDOWB or VWINDOWTL instruction is executed This is an array of switches one for each virtual camera See the general description of switches in AdeptVision User s Guide for syntax details V BACKLIGHT affects binary operations only and has no effect on grayscale processing Details If the VBACKLIGHT switch is enabled the vision system considers the visual background to be white Otherwise the background is considered to be black The objects to be recognized are in the foreground the opposite color of the background The borders of the image that is beyond the image limits set by the parameters V FIRST LINE V LAST LINE V FIRST COL and V LAST COL are always treated as background The state of V BACKLIGHT is critically important for prototype training for clear grip tests and for blob analysis You may train prototypes only on foreground objects Gripper positions over foreground regions are never clear VMIN AREA applies to outermost foreground regions V MIN HOLE AREA applies to all inner regions both foreground and background However the state of VIBACKLIGHT is not critical for prototype recognition Prototypes are either black or white In fact one prototype may be black while another is white and both may be found in the same image regardless of the stat
337. transition For an arc ruler the values are angular distances in degrees from the start to each transition For a graylevel type 1 ruler the values are pixel graylevel values along the ruler The graylevels are not valid if the ruler was clipped as described above indicated by 1 as the second value in the array Optional integer values that identify the first array elements to be defined in datal and mags respectively Zero is assumed for any array index that is omitted If a multiple dimension array is specified only the right most index is incremented as the values are assigned Optional array into which the signed edge magnitudes are placed for each of the edges returned in data For convenience the count and color clipped values are duplicated into the first two AdeptVision Reference Guide Rev A Program Instruction VRULERI elements defined in this array This array is valid only for type 2 and type 3 rulers ibr Integer value specifying the image buffer region for the ruler Image buffer regions specify both a size and a frame store see the description of VDEF AOD shape Real valued expression indicating how the ruler path is described 1 for line 2 for counter clockwise arc 3 for clockwise arc The default is 1 linear ruler x0 y0 Real valued expressions indicating the coordinates in millimeters of the starting point for the linear ruler len Real valued ex
338. tring variable specifying a template for definition a font for training or all fonts for planning The string naming a font has the form FONT_n where n is the number in the range 1 to 99 of a single font or FONT_0 specifies all fonts Ms yy The string name of a template has the form TMPL_n where n is the number of the template in the range 1 to 99 For OCR training only Optional string variable that specifies the characters in the font that are to be trained The text may contain duplicate characters Spaces are ignored Integer value specifying the image buffer region for training a model Image buffer regions specify both a size and a frame store see the description of VDEF AOI The image buffer region s AOI must specify a nonrotated rectangular shape For correlation templates the AOI is reduced if necessary to a multiple of four pixels Optional real valued expression indicating the shape of the window Currently the only choice is 1 for a rectangular window NOTE If any of the following four parameters cx cy width and height are specified all four parameters must be specified They are always needed for defining a template or training on a font but they are never needed for planning a font cx cy Real valued expressions specifying the center coordinate of the rectangular training window or template definition window in millimeters AdeptVision Reference Guide Re
339. truction VCOPY program instruction VDEETRANS program instruction VEDGE program instruction VGET AOI program instruction VGET TRANS program instruction VHISTOGRAM program instruction VMORPH program instruction VPICTURE program instruction VSUBTRACT program instruction VCORRELATE program instruction VFIND ARC program instruction VFIND LINE program instruction VFIND POINT program instruction VOCR program instruction VRULERI program instruction VTRAIN program instruction VTRAIN MODEL program instruction VWAIT program instruction VWINDOW program instruction VWINDOWB program instruction VWINDOWI program instruction AdeptVision Reference Guide Rev A 55 VDEFCONVOLVE Program Instruction Syntax VDEF CONVOLVE type array i j Function Define an image convolution Usage Considerations VDEF CONVOLVE is intended only for experienced users It is not critical to the operation of the vision system Convolutions larger than 3x3 require the AdeptVision Enhanced VXL Interface option When a convolution with a definition larger than 3x3 is performed by a standard vision system only the center 3x3 elements are used in the convolution Thus predefined convolutions 1 to 6 will perform correctly on all systems Convolutions 7 to 14 will run on all systems but may not return the expected results Parameters type Integer value in the range 17 to 32 inclusive This is the number of the user definable convo
340. truction to be executed Parameters proto String expression specifying the name of the prototype for which the grip is being defined or deleted grip Real valued expression with a value in the range 0 to 4 A gripper number of 0 deletes all defined grippers for the specified prototype If the gripper number is in the range 1 to 4 it indicates which gripper is being newly defined for the prototype mode Real valued expression interpreted as TRUE or FALSE indicating how invisible fingers are to be treated Fingers are invisible if they are outside the field of view If the mode value is FALSE zero invisible fingers are not considered clear Invisible fingers are assumed to be clear if the mode value is TRUE nonzero num_fngrs Real valued expression with a value in the range 1 to 5 specifying the number of rectangles that are used to model the gripper fingerprints trans Array of transformations that define the 2 D location and orientation of the rectangles The X Y and RZ components of each transformation define the center position and orientation of one rectangle relative to the position of the prototype num_fngrs number of transformations must have been stored in sequential array elements ip dy K These are optional array indexes that specify the first elements to be accessed in the respective arrays Zero is assumed for any array index that is omitted If a multiple dimension array is specified only the right most
341. ture and compute the histogram enter the V monitor commands VPICT cam 2 and VHIST where cam is the number of the virtual camera being used or use the mouse to make the menu selections to perform these same operations The goal is to have the video data fill most of the intensity range 0 to 127 without spilling over either end If the video data spills over the left end the histogram curve shows a spike over the 0 intensity label Similarly if the video data spills over the right end the curve shows a spike at or near the 127 intensity label You should increase V GAIN to expand the intensity range of the video data or decrease V GAIN to reduce the intensity range Similarly you should increase V OFFSET to shift the video data toward the left and decrease V OFFSET to shift it toward the right For good results you may have to repeat the procedure of taking a picture computing the new histogram and adjusting V GAIN and V OFFSET a few times V GAIN must be assigned an integer value in the range 1 to 256 inclusive The parameter is set to 128 when the V and AdeptVision systems are loaded into memory from disk 132 AdeptVision Reference Guide Rev A System Parameter Example Make the video gain 120 for all virtual cameras PARAMETER V GAIN 120 Related Keywords VHISTOGRAM monitor command and program instruction V OFFSET system parameter AdeptVision Reference Guide Rev A V GAIN
342. turn For an edge finding ruler this causes processing to stop after the specified number of edges are found For a graylevel type 1 ruler this specifies the exact number of pixel values to return starting from the start of the ruler The default is 1 to return as many values as possible Adept Vision Reference Guide Rev A 213 Program Instruction Optional real valued expression specifying the polarity of the edges to be considered by the ruler 1 Consider only light to dark edges 0 Consider all edges default 1 Consider only dark to light edges NOTE The parentheses in the instruction syntax can be omitted if all five of the above parameters are omitted VRULERI edge dir data i j mags 214 Real array into which the transitions detected by the ruler are placed The data format of this array is count color clipped value_1 value_2 value_n count is the number of transitions detected by the ruler The color clipped value is the starting color of a binary ruler ruler types 2 1 and 0 For a type 1 type 2 or type 3 ruler the value is 1 if any part of the ruler is clipped by an edge of the frame store Otherwise the value is 0 The values value_i represent the data returned by the ruler The interpretation of these values depends on the shape and type of ruler For a linear ruler the values are millimeter distances along the ruler from the start to each
343. types Optical Character Recognition OCR fonts or correlation templates Usage Considerations The VISION switch must be enabled and the vision processor must be idle for this command to be executed Parameters file spec model_name Details Specification of the disk file into which the models are to be stored This consists of an optional physical device an optional disk unit an optional directory path a filename and an optional file extension Uppercase or lowercase letters can be used The current default device unit and directory path are considered as appropriate see the DEFAULT command in the V Operating System User s Guide If no filename extension is specified the extension VS is appended to the name given Optional name of a current prototype OCR font or correlation template to be saved in the file Uppercase or lowercase letters can be used If no model names are listed all prototypes are stored In this case the equal sign can be omitted Font names have the form FONT_n where n is the number of the font in the range 1 to 99 The special font name FONT_0 refers to all the fonts Template names have the form TMPL_n where n is the number of the template in the range 1 to 99 The special template name TMPL_0 refers to all the templates The type of model being stored is determined by the name which has been trained as either a prototype or an ObjectFinder
344. uide Rev A Keyword Type KEYWORD Parameters The requirements for input and output parameters are explained in this section If a parameter is optional it will be noted here When an instruction line is entered optional parameters do not have to be specified and the system will assume a default Unspecified parameters at the end of an argument list can be ignored Unspecified parameters in the middle of an argument list must be accounted for by commas For example the following keyword has four parameters the first and third are used and the second and fourth are left unspecified SAMPLE INST var_1 test String and numeric input parameters can be constant values 3 32 part_1 etc or any legitimate variable name see Chapter 1 of the V Language User s Guide for the requirements of a variable name The data type of the constant or variable must agree with the type expected by the keyword String variables must be preceded by a Precision point variables must be preceded by a Belt variables must be preceded by a String constants must be enclosed in quotes Real and integer constants can be used without modification V keywords cannot be used as variable names see Appendix A for a complete list of keywords Details This section describes the function of the keyword in detail Examples Examples of correctly formed instruction lines are presented in this section Related Keywords Additional k
345. uld be set to 1 5 or larger In both cases if it is set to less than 0 5 the finders automatically use the value 0 5 The other function of VMAX PIXEL VAR applies during boundary analysis when a VWINDOW VPICTURE in mode 0 or 1 or VTRAIN operation is executed During boundary analysis the vision system characterizes the regions in the image as a connected sequence of lines and arcs The lines and arcs are an approximation of the boundary of the binary image V MAX PIXEL VAR determines the maximum number of pixels by which the lines and arcs may Adept Vision Reference Guide Rev A 167 V MAX PIXEL VAR System Parameter deviate from the binary image boundary When this tolerance is increased edges are smoothed and fewer lines and arcs are generated On the other hand a more accurate model of the boundary is attained when this tolerance is decreased For all practical purposes VIMAX PIXEL VAR should be kept in the range 1 0 to 3 0 If VMAX PIXEL VAR is 0 both line and arc fitting are disabled This may be useful for some applications that use the VEDGE INFO instruction and want only the primitive edges that bound regions that is the edges displayed when the V SHOW EDGES system switch is enabled The allowable range for this parameter is 0 0 to 8 0 inclusive The parameter is set to 1 5 when the V and AdeptVision systems are loaded into memory from disk Exa mple The following figure illustrates the effects on boundary analysis
346. us asynchronous and asynchronous with reset If VSYNC STROBE is set to 2 strobing is asynchronous and the camera vertical drive is reset If VSYNC STROBE is set to 1 strobing is synchronous and image quality is highest If VSYNCH STROBE is set to 0 strobing is asynchronous and response time is shortest Only the values 0 1 or 2 can be assigned to VSYNC STROBE The default setting is 0 Normally the cameras used with AdeptVision VXL comply with the interlaced RS 170 video standard Every 60th of a second one field even or odd is read from the camera A full frame consists of two fields an even field and an odd field When the value of the system parameter V ISYNC STROBE is 2 the vertical drive will be reset when a picture is taken The strobe signal will be output at the same line as for synchronous strobe V SYNC STROBE of 0 This line is defined in the camera model as delay_strobe Adept Vision Reference Guide Rev A 257 V SYNC STROBE System Parameter CAUTION Only some cameras have this feature And there are always some internal or external switches that need to be set 258 properly for this mode to operate See your camera s manual for details Camera model 2 supports this mode Camera models 3 4 and 5 MUST have this mode selected When V SYNC STROBE is set to 0 the strobe light is fired asynchronously as soon as a VPICTURE request occurs However for interline transfer CCD cameras th
347. useful for inspection or for implementing unusual recognition strategies As an example of inspection components missing from printed circuit boards may be discovered by backlighting the boards and looking for unexpected holes The memory used for storage of the hole information is taken away from the memory normally allocated for the object queue The object queue holds the objects seen until VLOCATE instructions are processed Every satisfied VLOCATE instruction removes one object from the object queue leaving room for a new object The object queue is displayed when the VQUEUE monitor command is issued The vision system queue has a maximum capacity of 1200 objects and holes Each hole in the queue takes the place of one object Related Keywords 148 VFEATURE real valued function VLOCATE program instruction V DISJOINT system switch AdeptVision Reference Guide Rev A System Parameter V IO WAIT Syntax V IO WAIT camera Function Enable the synchronization of taking pictures VPICTUREs with an external event that triggers the fast digital input interrupt line Usage Considerations Operation of the external trigger can be configured with the Adept controller configuration program in the file CONFIG_C V2 on the Utility Disk See the V Language User s Guide for details on digital I O A change to this parameter takes effect when the next VPICTURE command or instruction is executed This is an array of parame
348. v A Program Instruction VTRAIN MODEL dx Real valued expression specifying the width of the window in millimeters dy Real valued expression specifying the height of the window in millimeters ang Optional real valued expression specifying the orientation of the OCR training window in degrees The default is 0 degrees When defining a correlation template this must be omitted or have the value 0 Details This instruction is used for both correlation templates and OCR fonts Conelation Templates VTRAIN MODEL defines and fully trains a correlation template in one step This is simple compared to prototype or font models which require a definition step followed by multiple training steps with different example images A correlation template is merely a grayscale window extracted from the image VTRAIN MODEL analyzes the template and determines the default depth for hierarchical search and skip patterns for each level of hierarchy The new information is stored in the template The position and size of the window is specified with the VTRAIN MODEL instruction When VTRAIN MODEL is executed the pixels within the window are copied from the frame store into the vision CPU memory It remains there until it is deleted via VDELETE NOTE Template widths must be a multiple of 4 pixels and will be reduced if they are not an exact multiple You should consider the memory requirements of templates when using correlation in application
349. vailable only with the AdeptVision Enhanced VXL Interface option The statistics returned are for all bounded regions within the area of interest The data array returns the following information data i Is the area of interest clipped by the field of view 1 yes 0 no Area of blobs in pixels X component of the blob centroid in millimeters Y component of the blob centroid in millimeters X component of closest point on blob perimeter in millimeters X component of furthest point on blob perimeter in millimeters Y component of closest point on blob perimeter in millimeters Y component of furthest point on blob perimeter in millimeters Perimeter of all bounded areas type 2 window only in millimeters data i 1 data i 2 data i 3 data i 4 data i 5 data i 6 data i 7 data i 8 RA A A A AdeptVision Reference Guide Rev A Program Instruction VWINDOWB Related Keywords VDEEFAOI program instruction VPICTURE program instruction VWINDOWI program instruction AdeptVision Reference Guide Rev A 289 VWINDOWI Program Instruction Syntax VWINDOWI cam type dmode sample data index ibr VWINDOWI cam type dmode sample data index shape cx cy dx dy ang VWINDOWI cam type dmode sample data index shape cx Cy Or ir ang0 angn Function Extract image information from within a wndow with any of the following shapes rectangle circle pie cut ring or ring segment
350. vel 0 25 Convergence measure 26 samples taught integer 27 features in the model integer 28 Max pixel variance 29 Max loc distance Indexes 12 19 and 20 are a set that is computed following multi instance training See the descriptions following Table 2 1 for each item in Table 2 3 Exceptions are shown below ITEM 12 Minimum verify percentage is a recognition certainty value The verify percentage following a VSHOW instruction is the minimum verify percentage required for recognition of an instance of the finder model ITEM 17 Number of pairs Following VSHOW this is the number of pairs in the finder model ITEM 19 Average verify percentage is a recognition certainty value When following a VSHOW instruction this is the average verify percentage required for recognition of an instance of the finder model ITEM 20 Max verify percentage is a recognition certainty value When following a VSHOW instruction this is the maximum verify percentage required for recognition of an instance of the finder model ITEM 21 Hierarchical level This is the image processing level 0 2 See VTRAIN FINDER for details ITEM 25 Convergence measure This is a measure of the convergence of the feature weights during multi instance training Higher means more convergence i e higher confidence that the weights are really converging to some set of constant values This would mean that the model is stabilizing In most
351. ver a 6x6 neighborhood Use of the sampling feature can greatly reduce the amount of memory needed to store an image A sample rate of 2 results in an image that is 1 4 the size of the original Sampling may be used to save several reduced images for simultaneous display see the description of VPUTPIC later in this chapter Normally you never have to examine or alter the header string However for completeness its contents are described in Table 2 4 below The VPUTPIC instruction requires that this information be present and consistent with the picture data stored in the rest of the array Therefore you should not change the header string without a full understanding of the format The items in the header string are all integer values They may be extracted using the V INTB function Table 2 4 Contents of VGETPIC VPUTPIC Header String Start Char Item Description 1 version Version number for maintaining compatibility 3 system Type of system 2 for area grayscale systems 5 camera Virtual camera accessed by VGETPIC 7 type Type of image data 1 2 3 4 pre 11 0 images only 9 packed 0 for 1 pixel per byte 1 for 8 pixels per byte 11 sample_rate Sample rate used 13 sample_mode 0 for sampling 1 for averaging 15 shape 1 for rectangular 17 x0 Actual x0 of starting location 19 y0 Actual y0 of starting location 21 dx Actual number of image columns read and stored 23 dy Actual number of imag
352. w with any of the following shapes rectangle circle pie cut ring or ring segment Adept Vision Reference Guide Rev A VWINDOW 299 Prototype Recognition Algonthms Overview ww aa a a a 302 Connectivity Analysis aa aaa a a a 302 Chain Encode Per meters 303 Ft Primitive Edges to Chains 0 0 0 303 Ft Lines and Arcs to EdgeS 058 303 Classify Features ddd aaa a a 304 Propose Prototype to image Matches 304 Verify Match 0 305 AdeptVision Reference Guide Rev A 301 Appendix B Overview Overview When AdeptVision VXL processes an image doing boundary analysis and recognition the operations on the image data are performed in sequence The camera image is frame grabbed and stored in memory Then connectivity is performed on the binary possibly thresholded edges The boundaries of the regions found are then processed Image processing involves multiple steps producing multiple boundary representations the last of which is a connected sequence of lines and arcs Finally recognition is performed by comparing boundary features with the features of the prototype models More specifically there are seven steps to image processing Connectivity analysis Chain encode perimeters Fit primitive edges to chains Fit lines and arcs to edges Classify features Propose prototype to image matches Se oT oS S Do Verify match
353. width angle Function In a search window find the edge point that is nearest to one side of the window Usage Considerations The frame store currently selected must contain a valid picture Otherwise an error results If the calibration being used includes correction for perspective distortion the correction is applied to the found point returned Adept recommends that you use the first syntax Parameters cam Optional real valued expression that specifies a virtual camera number The default is 1 The camera number is used to pick the set of switches and parameters to use pos Optional real valued expression indicating the starting point in the search window 1 dark side 1 light side The default value of pos is 1 dmode Optional real valued expression specifying the display mode for this operator The choices are 1 no draw 0 erase 1 draw solid 2 complement 3 draw dashed and 4 complement dashed The default is 1 draw solid effort Optional real valued expression indicating the effort level to use when searching for edge points The effort level can range from 1 least effort to 100 maximum effort The default is 100 type Optional real valued expression specifying the type of point finder 2 dynamic binary 1 raw binary 0 run length binary 2 fine edge The default is 2 fine edge data Real array describing the outcome of the search as follows data i 0 TRUE if a
354. ws See below avg Average graylevel in the window min max The minimum and maximum graylevel in the window object bkgd These are the numbers of pixels that are part of an object or part of the background bkgd stdev The standard deviation of the graylevels in the window from the mean avg ecount The number of edge points in the window A type 6 window returns histogram data in its array parameter The array is filled with the pixel counts for each of the possible intensity values 0 to 127 292 AdeptVision Reference Guide Rev A Program Instruction VWINDOWI All the other types of windows return the boolean clipped indicating whether or not the window was completely in the image Nonrectangular windows circles rings etc that are clipped are not evaluated That is zeros are returned in the array elements for all the items other than clipped Rectangular windows that are clipped are still evaluated but the fact that part of the window was out of the image could invalidate usage of the window information All windows except type 6 windows also return their width height and area in pixels For windows orthogonal to the image boundaries the area is simply width times height If the sample parameter is 2 however the area is approximately width height 4 You may notice that a window defined to be square in millimeters does not have equal width and height in pixels This happens because the pixel
355. x VGETCAL cam Function Program Instruction VGETCAL scalers i pmm to pix j k pix to pmm 1 m to cam Ask the system to fill in arrays with the previously defined vision calibration data for a given virtual camera Parameters cam scalers pmm to pix pix to pmm j k 1 m to cam Details Optional real valued expression that specifies the virtual camera number Real array that receives the scaler calibration values Optional two dimensional real array that receives the millimeter to pixel transformation matrix If this array is specified the array pix to pmm must also be specified Optional two dimensional real array that receives the pixel to millimeter transformation matrix If this array is specified the array pmm to pix must also be specified Optional integer values that identify the first array element to be defined in the respective array Zero is assumed for each index that is omitted If an array is specified that has more dimensions than needed only the right most indexes are incremented as the values are assigned Optional variable that recieves the vision transformation This instruction is used by the calibration programs supplied with the vision system The calibration programs perform a VGETCAL to determine what calibration data was previously defined via VPUTCAL On power up the vision system preassigns calibration data for all virtual cameras Virtual
356. xpression indicating whether or not to wait for the image acquisition to complete The choices are 1 to wait for completion of the image acquire and the first stage of processing run length encoding 0 to wait only for the image acquire to start 1 for no waiting at all Values of 0 and 1 apply only to quick frame grabs mode 2 see below The default is 1 wait for the completion Optional real valued expression that specifies the grayscale frame store into which the image is acquired Currently only the virtual frame buffer referenced in the image buffer region is used The full frame is used regardless of any AOI defined for the image buffer region The default is 1011 Optional real valued expression that specifies the grayscale frame store to be selected for subsequent processing Currently only the virtual frame buffer referenced in the image buffer region is used AdeptVision Reference Guide Rev A Monitor Command and Program Instruction VPICTURE The full frame is used regardless of any AOI defined for the image buffer region The default is the value of acq_ibr indicating the frame buffer into which the image is acquired This parameter applies only to quick frame grabs mode 2 In other modes the selected frame is made equal to acq_ibr NOTE The parentheses in the instruction syntax can be omitted if all four of the above parameters are omitted mode how_many Details Real valued expressio
357. y be trained in different orientations such as first horizontally then diagonally The system automatically rotates the character features so that the given window orientation matches the base orientation of the font All characters in a font share the same base orientation The base orientation is defined by the AOI used during the first training instance with the font Thus if you first trained A using an AOI with the angle 30 degrees and then trained B using an AOI with the angle 0 the B would be rotated 30 degrees to bring it into alignment with the font s base orientation of 30 In addition to training VTRAIN MODEL may be used to make the vision system plan a font for recognition This is not a crucial step however because the VOCR instruction automatically plans a font that has not been planned However since planning may take a few minutes you may prefer to control when that occurs by executing VTRAIN MODEL in an initialization program When a font is planned and its display is enabled plan 1 the characters in the font are displayed at the top of the Vision window as they are planned Then a discrimination matrix is displayed so that you can watch the effects of the vision system s auto weighting of features The color coding in the matrix indicates how the average character horizontally scores with respect to the model vertically The colors and associated score ranges are listed

AdeptVision Reference Guide, Rev A

Contents

Download Pdf Manuals

Related Search

Related Contents