as PDF
Contents
1. as the software is able to avoid it from the measure but this is not possible with standard instruments and the influence of the burr is different depending on the operator that is measuring So it is possible to say that measurement process doesn t change its precision passing from 1 to 2 operators vice versa manually measuring it loose precision and repeatability as higher is the number of operator It means also that Cp and Cpk decrease and the process could results out of control In fig 18 are reported system performances about reproducibility for each feature Finally once the data has been acquired more and more quality control instruments can be automatically implemented as example quality control chart D C Montgomery 1985 About time saving before explain the results it is proper to describe the operations carried out both in manual and in optical controls In manual control the operator have to execute the following steps in the order Prepare sheet for registration Control screw termination by a microscope Measure burr entities by a contour projector Measure feature D1 D2 and D4 Execute Functionality test try to screw the coupled particular Measure feature D3 with go no go gauge The order of the operation is important because the entity of some feature can be changed during control For example go no go gauge control can damage the particular and functionality test can be compromised as also screw2. s also possible to increment illumination intensity but in this case the light passes through the material in a huge way and the dimension of the part will be undervalued so this method have to be used carefully This process have to be done confronting image results with contact methods or other suitable techniques in particular standard caliper measurement was used here 3 1 Simple feature measurement In this section measurement process of simple feature will be illustrated A simple feature is a feature that requires the identification of two elements then some dimensional or geometric information are extracted by them Features number 1 3 and 4 of fig 6 b are representative of this case Feature number 1 will be treated as example The objective is to determine the diameter of the part As previously explained two search areas have to be determined The region may have the desired shape circular rectangular square polygonal depending on the feature to inspect In this case because a line is searched two rectangular ROI are defined green rectangles in fig 8 b The ROI has fixed size and fixed position in the image hence alignment process is important in order to bring the part always in the same position in the image permitting the software to find the edge at all times with high reliability and robustness A rake division of five pixels has been used distance between each inspection line it means that a point every 0 085 mm
3. C 1985 Introduction to statistical quality control j Wiley amp sons 0 471 80870 9 New York National Instruments 2005 Labview user manual National Instruments 2005 NI Vision concept manual Russ J C 1994 The image processing handbook second edition CRC Press 0 8493 2516 1 United States of America www intechopen com Applications and Experiences of Quality Control QUALITY CONTROL Edited by Prof Ognyan Ivanov bitei ty ryan heare ISBN 978 953 307 236 4 Hard cover 704 pages Publisher InTech alg wi F o a GUARANTEE nd Ped Published online 26 April 2011 Published in print edition April 2011 The rich palette of topics set out in this book provides a sufficiently broad overview of the developments in the field of quality control By providing detailed information on various aspects of quality control this book can serve as a basis for starting interdisciplinary cooperation which has increasingly become an integral part of scientific and applied research How to reference In order to correctly reference this scholarly work feel free to copy and paste the following Marco Sasso Massimo Natalini and Dario Amodio 2011 Digital Image Processing for Quality Control on Injection Molding Products Applications and Experiences of Quality Control Prof Ognyan Ivanov Ed ISBN 978 953 307 236 4 InTech Available from http www intechopen com books applications and experiences of quality control digit
4. a different procedure that uses threshold method has been applied Now the problem is to measure correctly the free area for fluid flow With standard methods is only possible to have an approximation measuring the burr height and computing the area of the annular region occupied by burr But if this area is not annular or presents an irregular form then it is almost impossible to get a precise measurement of its extension Using digital images it is possible to implement a series of operations that compute the area with a good precision and automatically www intechopen com Digital Image Processing for Quality Control on Injection Molding Products 567 Since the free area has the highest gray level of the image the software developed computes the mass centre of the image substituting mass with pixel gray level values obviously which will always fall in the free area From this point the software begins to select all the pixels having a gray level comprised in a properly defined range Since a free area is looked for a value of 255 has been set with a tolerance of 5 gray level The algorithm stops at the burr edge when it find a gray level lower than 250 At the end of selection process the software counts the number of pixel selected and applying calibration information determines the area available for fluid flow In fig 12 a is reported the image before processing and fig 13 b shows the processed image a b Fig 13 The proce
5. gauges but it s impossible to get an exact value of the diameter The use of a standard gauge is difficult because there are not plane surfaces Digital images offers a good solution It is possible to determine a circular edge with the same method explained before having care of changing the search area which now has to be annular green circles in fig 9 c with inspection lines defined by their angular pitch along the circumference white arrows in fig 9 c The result is the red circle reported again in fig 9 c which is the circle that best fits the detected circumference points 3 2 Composed feature measurement In this section the measurement of feature 2 fig 5 b will be explained briefly as it can be carried out by a simple extension of the procedures already developed Now the aim is to determine the distance between the ribbing edge and the axis of the hollow cylinder identified before and whose diameter has been measured So the feature involved are three The procedure can be similar to the previous starting from the cylinder edges already determined see fig 8 b the equation of their medium axis can be easily calculated the axis is reported in fig 10 a Also the ribbing edge is determined with the procedure illustrated before red in fig 10 a and the distance between the axis and the rib edge can be evaluated easily and quickly by analytical computations The same method can be applied in all that situation when it s
6. identify and measure the main dimensional features of the components such as overall size edges length hole diameters and so on Of particular interests is the use of digital images for evaluation of complex shapes and dimension where the distances to be measured are function of a combination of several geometric entities making the use of standard instrument as callipers not possible At the same time methods used for image processing will be presented Moreover a description of system performances related to quality product requirements will be presented Second application examines the possibility to identify and quantify the entity of burrs The case of a cylindrical batcher for soap in witch its effective cross sectional area have to be measured will be showed together with a case study in witch burrs presence could bring to incorrect assembly or functionality of the component Threshold and image subtraction techniques used in this application will be illustrated together with the big number of information useful to manage production process Third it will be presented a possible solution to the problem of identifying general shape defects caused by lacks of filling or anomalous shrinkage Two different approaches will be www intechopen com 556 Applications and Experiences of Quality Control used the former that quantifies the general matching of whole images and the latter that inspects smaller areas of interest Finally a
7. is stored about 40 points per line When points are determined they are used to fit a line with a linear regression method and to determine the equation of the line in the image Once the equation has been determined the process is repeated for the second edge to determine the second equation The rectangular search area is the same and has the same x position this is important to define distance in following steps Now using the equations is possible to find all the information required analytically Defining first line as reference line maximum and minimum distance between the lines has www intechopen com 564 Applications and Experiences of Quality Control been computed the distance between extreme points and has been averaged to find medium distance between edge also parallelism error can be evaluated as the difference between maximum and minimum edge distances For each piece two information have been obtained medium distance and geometric error The edge determination is applicable also to other feature such as holes indeed in the following case study the problem of a threaded hole is treated where the inner diameter has to be measured a Fig 9 A problem of inner diameter determination Traditional methods used for control process employs go no go gauges for diameter verification So it is possible to say that inner diameter is comprised between established tolerance range minimum and maximum dimension of
8. maximum height of burr and it is inadequate www intechopen com 566 Applications and Experiences of Quality Control to determine precisely the free area of the hole or the extension on the edge Digital images solve the problem using a combination of threshold and binary operation so these two techniques will be now explained In a few words in the digital images here presented parts are characterised by an intensity range of gray level much different from the background Pixels of the part have a gray level intensity comprised in a given interval while all other pixels not included in this intensity range can be considered as background Threshold operation sets all pixel that belong to the desired interval to a user defined value typically 1 and all other pixel of the image to zero The image obtained is a binary image First case presents a burr that has a gray level between the free area gray level 250 and the part gray level lt 120 The software after alignment operation extracts the zone of the image interested by analysis fig 12 b then it performs a first binarization on this sub image with a threshold value of 250 settings to 255 all pixels with a gray level lower than 250 in order to obtain the area of the part and the burr together fig 12 c To the same image extracted a new binarization is applied with a gray level value of 138 and all pixel with a lower gray level value are set to 255 to obtain the area of th
9. of small black dots All these application are based on histograms obtained by the image The histogram consists in counting the total number of pixel at each level of the greyscale The graph of an histogram is a very quick way to distinguish if in the image there are regions with different grey level value Histograms are useful also to adjust the image and to optimize image acquisition conditions www intechopen com Digital Image Processing for Quality Control on Injection Molding Products 575 The histogram is the function H defined on the grayscale range 0 k 255 such that the number of pixels equal to the grey level value kis NI Vision Concept Manual 2005 H k n 10 where k is the gray level value nx is the number of pixels in an image with a gray level value equal to k n Xn from k 0 to 255 is the total number of pixels in an image The histogram plot reveals easily which gray level occurs frequently and which occurs rarely Two kind of histogram can be calculated linear and cumulative histograms In both cases the horizontal axis represents the gray level value that ranges from 0 to 255 For linear histogram vertical axis represents the number of pixels nx set to the value k In this case the density function is simply given by 10 The probability density function is Gonzalez amp Woods 1992 Pine k n 0 11 where Prinea k is the probability that a pixel is equal to k For cumulative histog
10. termination can be damaged and incorrect consideration can be done by the operator DIR ONE www intechopen com Digital Image Processing for Quality Control on Injection Molding Products 573 35 Reproducibility Feature D1 30 Dftware 25 Operators Tolerance Limits 20 13 600 13 700 13 800 13 900 14 000 14 100 14 200 14 300 a Reproducibility feature D1 Reproducibility Feature D3 25 20 Oftware Tolerance Limits 4 400 4 500 4 600 4 700 4 800 4 900 5 000 5 100 c Reproducibility feature D3 Fig 18 Reproducibility of the measurements 35 Reproducibility Feature D2 30 25 Software 20 Operators Tolerance limits 19 100 19 200 19 300 19 400 19 500 19 600 19 700 19 800 b Reproducibility feature D2 Reproducibility Feature D4 Sftware Operators Tolerance Limits 16 700 16 800 16 900 17 000 17 100 17 200 17 300 17 400 d Reproducibility feature D4 Functionality E D4 D3 m D2 m Di m Burr Control E amp rew Termination m heet preparation __FEATUREINSPECTED 0 1 TIME min i i a Time needed for manual control FEATUREINSPECTED Position Change E Vertical Pos Orizontal Pos Functionality FEATURE NSPECTED TIME min b Time needed for automatic control Automatic Manual TIME min c Differe
11. 1 18 60 1 00 a b Fig 5 a the piece to analyse b requested features More techniques are based on use of pixel mask that runs through the image computing the sum of products of coefficients of pixel mask with the gray level contained in the region encompassed by the mask The response of the mask at any point of the image is Gonzales amp Woods 1992 9 R WZ WZ Woe gt WZ 2 i l where wi coefficient of the pixel mask Zi gray intensity level of the pixel overlapped Using different kind of mask different features can be detected All of them are detected when IR gt T 3 where T is a nonnegative threshold Fig 6 shows different masks for detection of different features b Horizontal edge a Point c Vertical edge d 45 edge Fig 6 Different pixel mask for different features In this application an easier method based on pixel value analysis along a pixel line has been used It is a simplification of derivative operators method Gonzales amp Woods 1992 that uses gradient operators and analyses gradient vector to determine module and direction of the vector www intechopen com 562 Applications and Experiences of Quality Control Usually an edge is defined as a quick change in pixel intensity values that represents boundaries of an object in the FOV It can be defined by four parameters 1 Edge strength defines the minimum difference in the greyscale
12. 28 Digital Image Processing for Quality Control on Injection Molding Products Marco Sasso Massimo Natalini and Dario Amodio Universita Politecnica delle Marche Italy 1 Introduction The need to increase quality of products forces manufacturers to increase the level of control on finished and semi finished parts both qualitatively and quantitatively The adoption of optical systems based on digital image processing is an effective instrument not only for increasing repeatability and reliability of controls but also for obtaining a large number of information that help the easily management of production processes Furthermore the use of this technology may reduce considerably the total amount of time needed for quality control increasing at the same time the number of inspected components when image acquisition and post processing are feasible in real time the whole production can be controlled In this chapter we describe some quality control experiences carried out by means of a cheap but versatile optical system designed and realized for non contact inspection of injection moulded parts First system architecture both hardware and software will be showed describing components characteristics and software procedures that will be used in all the following applications such as calibration image alignment and parameter setting Then some case studies of dimensional control will be presented The aim of this application is to
13. 5 60 e 60 65 f 65 70 Fig 22 Six groups of different intensity colours 7 2 Colour uniformity determination The second application developed regards the mixing problems between virgin polymer and masterbatch colour In this case components appears with non uniform colour as showed in figure 23 Figure 23 a shows a particular in which white material hasn t been correctly mixed while the component of Figure 23 b was obtained with a correct mixing with uniform rose Figures 23 c and 23 d show the acquired images of the same particulars with the system Extracting the same area from the images as illustrated in fig 20b the histograms of figure 23e are obtained it s evident that histogram of first particular has a standard deviation bigger than the second So it is possible to establish a tolerance range in which the variation can be accepted while outside the parts have to be discharged m NON UNIFORM m UNIFORM 90 100 110 120 130 Fig 23 Example of non uniform colour 7 3 Detection of small black dots This problem is generally due to small impurities included in the material which appear as small dots in the final product Often this kind of control is still demanded to visual inspection of expert operators who look at the particulars from a distance of about 40 cm www intechopen com 5 8 Applications and Experiences of Quality Control and if they can see the black dots the pieces are rejected In th
14. al image processing for quality control on injection molding products INTECH open science open minds InTech Europe InTech China University Campus STeP Ri Unit 405 Office Block Hotel Equatorial Shanghai Slavka Krautzeka 83 A No 65 Yan An Road West Shanghai 200040 China 51000 Rijeka Croatia AB bhAa Be RBA EDA R405 470 Phone 385 51 770 447 Phone 86 21 62489820 Fax 385 51 686 166 Fax 86 21 62489821 www intechopen com
15. aster Image can be now processed a b c d Fig 4 The alignment process applied to the first case of study 3 Dimensional measurement This section describes some examples of dimensional measurement on injection moulded parts In fig 5 a it is shown a 3D simplified model of the part to analyse It is constituted by two coaxial cylinders with a ribbing on the left side the material is Polyamide 66 Zytel 101 In fig 5 b all the features that have to be measured in quality control process are represented It s important to notice that there are two kind of feature to measure The dimensions denoted by numbers 1 3 and 4 require the identification of two simple feature edges of the part so this will be a simple feature to measure and a comparison with standard instrument is possible Instead feature number 2 requires the identification of an axis by identification of two edges and the identification of third element edge of the ribbing to measure the distance from the latter to the previous axis So three features are required and is impossible to get this measurement with standard instruments like gauges or callipers This represent a second kind of measurement that will be called composed feature Both cases pass through the identification of edges so the procedure used for their detection will be now described www intechopen com Digital Image Processing for Quality Control on Injection Molding Products 56
16. ction performances of the system in terms of measurement precision and time saving will be showed As example the part of fig 9 a has been used In fig 17 all the features to be determined are illustrated There are four dimensional measurements indicated as D a burr determination indicated as Burr and a shape defect identification indicated by screw end The description starts with dimensional measurement performance Table 1 shows nominal dimension of requested features and respective acceptable ranges To understand if the system is suitable to measure the required features and has a proper precision repeatability and reproducibility 100 images have been acquired from a component used as template moving each time the part in the FOV to simulate a real measurement process and resolution and standard deviation of results have been evaluated and compared with tolerance ranges About resolution the system uses an image in which a pixel corresponds to 0 016 mm while the minimum tolerance range is 0 15 mm so being their ratio approximately 10 the resolution can be considered suitable for the application www intechopen com 5 0 Applications and Experiences of Quality Control Screw end Burr Fig 17 The second component and the features to measure Nominal Dimension Acceptable range D1 400mm _ l3 82 14 18 mm Burr 20mm crew end Table 1 Feature to inspect Abou
17. e edge finish Starting from edge finish the first point where the greyscale value exceeds 90 of starting greyscale value is set as edge position Figure 7 b show the determination process of the edge and fig 7 a shows the edge position determinate yellow points O iT 1 Pixels 3 width 5 Contrast 2 Greyscale Values 4 Steepness 6 Edge Location ROI used for edge detection Edge position determination Fig 7 Determination process of edge position It s clear that parameter settings influence edge position especially when translucent material are analysed In fact translucent materials are characterised by lower greyscale intensity changes and high steepness values If external noise cannot be eliminated appropriately it could be difficult to get high repeatability in edge positioning in the image www intechopen com Digital Image Processing for Quality Control on Injection Molding Products 563 This can be avoided applying filters to the image before edge determination process In this application a filter to increment contrast was applied This allow to reduce the steepness to two pixels only and the filter width also to four pixels only In fig 8 filter effects on the image are shown fig 8 a represents the original image acquired fig 8 b reports filtered image with increased contrast pity a b Fig 8 a original image b aligned and filtered image To increment contrast in the image it
18. e of pieces could exceed the tolerance limits It s important to say that presented data are about 8 different cavities of the mould perhaps dividing data per each cavity only 1 cavity could present low Cp and Cpk indexes Modifying dimensions of that cavity indexes of the entire process could considerably increase Software Operator Software Operator Software rs ee es E Software Operator 2 5176 Table 2 Process performances Operator J J www intechopen com 572 Applications and Experiences of Quality Control Also same consideration about reproducibility of the measurements are possible To do this two operators measured the same sample of pieces once with the system and once with standard instruments Table 3 reports these results Software 48570 8540 Sd Dv 00197 ooa 0 0276 Table 3 Reproducibility of measurements It is possible to see that nothing is changed Negligible differences are there between average values and standard deviation in both cases and consideration about process performances are also valid On the contrary measured values normally change passing from 1 to 2 operators with manual measurements especially in feature D4 where is possible to notice a significant difference in average values about 0 03 mm Moreover standard deviation increases in all cases especially in D4 passing from 0 0230 to 0 0397 This is due to burr presence on the pieces see fig 11 a
19. e part only fig 12 d Finally image in fig 12 d is subtracted by fig 12 c to separate the burr areas Now with procedures similar to those used in edge detection it is possible to determine search areas in which ROI lines are defined to identify edges In this application an edge every 5 pixel has been determined and measured and the software extracts the maximum value within them and returns it as measurement result In this way only and always the maximum burrs height is considered in measurement results This method is conservative but gives the certainty that the defects entity will not exceed the part requirement limits a b c d e Fig 12 Burr determination process 1 HI H ii It s important to underline that is possible to obtain other information on burrs that cannot be extracted with traditional methods For example one can measure the area this procedure will be explained later to understand the entity of the problem a thin burr along all the edge indicates a defect solvable by proper tuning of injection process parameters while big area localized on a small part of the edge indicates a damage on the mould Furthermore it is also possible to determine x and y mass centre position to understand where the mould is damaged So much more parameters are available to control production process and it is easier to define a set of them that indicates the need to repair the mould In the case of fig 11 b
20. e time using the same instrument So as example measure features D1 D2 D4 and screw termination vertical position that requires three different instruments in a time of about 3 minutes and a half can be measured with only one instrument in only 1 minute Second the reduction of time amount also depends on the net time reduction in single feature inspection For instance burr determination in manual process requires over than 5 minutes while in automatic process requires only 1 minutes and a half together with feature D4 Finally time for sheet preparation and registration has been deleted because data are automatically stored and organized in a report that can be printed if necessary To close this section some limits of the system will be listed First of all only particulars or features with dimension smaller than the maximum FOV can be inspected So larger components have to be controlled with a different system Second precision of the system depends of FOV dimension In fact as bigger is FOV as lower is precision of measurement because of the number of pixel of camera sensor is constant Third only feature that lies in a plane or with a depth lower than depth of focus can be inspected 7 Aesthetic applications Digital image processing can be used also for aesthetic control Possible application with the developed system are three 1 Intensity colour determination 2 Uniformity colour determination 3 Detection
21. ew image using an algorithm of cross correlation so is possible to consider the template as a sub image T x vy of size K x L in a bigger image f x y of size M x N see fig 4 and the correlation between T and fat the pixel iJ is given by J C Russ 1994 XS T x y f x iy j E A E 1 ee PUP etiy s UUT xy Correlation procedure is illustrated by fig 3 Correlation is the process of moving the template T x v around the image area and computing the value C in that area This involves multiply each pixel of the template by the image pixel that it overlaps and then summing the results over all the pixels of the template The maximum value of C indicates the position where T best matches f a 0 0 a SO j i af T x fey Fig 3 Correlation process This method requires a big number of multiplications but is possible to implement a faster procedure first the correlation is computed only on some pixels extracted from the template to determines a rough position of the template then the correlation over all pixels of the template can be executed in a limited area of the entire image allowing to reduce processing time www intechopen com 560 Applications and Experiences of Quality Control In fig 4 the method applied to the first case of study is showed Fig 4 a reports the master image from which the template fig 4b has been extracted It s clear that only a part of the ma
22. g 15 b indicated by green arrow and the horizontal distance indicated by P between edge centre and thread centre is determined fig 15 c This parameter has been selected because of its great sensitivity with respect to even small damages of the thread Now is also possible to set a tolerance range within which the part can be accepted a Iiii 1 I I a b c Fig 15 Measurement process of damaged thread Here the maximum distance accepted is 1 mm which has been chosen analyzing more than 100 good pieces and assuming the greater dimension as the maximum tolerance limit Over this limit the software alerts the operator with a warning message suggesting to reject the part It s proper to note that expert operators are not required because now the measurement process is totally automated and the interpretation is not subjective however this theme will be discussed more deeply in next section Next case presented below is about individuation of process defects To do this part in fig 8 a will be considered In moulding process this particular often presents problems as short shot fig 16 a absence of ribbing and air bubbles on the ribbing and on the cylinder fig 16 b and 16 c respectively Different inspection methods are used depending on what is the defect looked for In fact for short shot defect that is generally easy to detect a simple pattern matching could be enough In this case an image of the en
23. get information about certain features and then narrows the FOV and acquire images with higher spatial resolution for capturing smaller details in the observed object Each position utilised has been calibrated sooner When a position is called the software restores related parameters and pass them to the following step for elaboration The software also control light intensity in the same way of previous components So all the information are passed to the acquisition step and then stored in the acquired images After this overview of the system it s proper to describe two functions that are used in all applications before any other operation image calibration and image alignment 2 1 Image calibration Using digital images two kind of calibration are necessary spatial calibration always and illumination and acquisition parameters calibration depending on material and shape of the pieces to be analysed Spatial calibration convert pixel dimensions into real world quantities and is important when accurate measurements are required For the application described below a perspective calibration method was used for spatial calibration NI Vision concept manual 2005 So the calibration software requires a grid of dots with known positions in image and in real world The software uses the image of the grid and the distance between points in real world to generate a mapping function that translates the pixel coordinates of dots into
24. inal part contacts the fixed plate of the mould interested zones are highlighted by red lines in fig 14 a These parts of the core are very thin and because of their low resistance property they are the parts of the mould to undergo damage in production process earlier If this part of the core is broken the injection moulded component will have a defected shape that not allows the coupled screw to slide in him Fig 14 b shows a correct thread termination while fig 14 c shows a damaged thread termination The difference is quite small and difficult to see for a human eye that is performing many checks repeatedly Nevertheless carrying out a quantitative measure of www intechopen com QI 68 Applications and Experiences of Quality Control such damage is basically impossible with standard instruments Instead using digital images it is possible to establish a procedure to determine a quantitative measure of the defect that is a number which should be compared with admissible tolerance limits SS F Naan ii a b c Fig 14 a the particular b Good thread termination c Damaged thread termination First step is always constituted by image alignment In this case the part is aligned with image reference system using external edges of the part then the inner diameter of the thread is determined together with his centre fig 15 a the thread termination edge is determined green in fi
25. is way the control totally depends on the operator judgment and is totally arbitrary The optical system here presented uses an algorithm based on particle analysis firstly it performs a binarization finding all pixels with an intensity value lower than a fixed threshold then it considers all the particles agglomerates of pixels and counts the number of pixels each particle is made of So by knowledge of the pixel size in real dimensions it is yet possible to establish maximum dimension of acceptable black dots and the control will not depend from the operators estimation In fig 24 a cap with black dots of different dimensions is reported In this case the software looks for object with dimension bigger than 4 pixels Dots with dimension bigger than the threshold value are highlighted in red circles while black dots smaller than four pixels are highlighted in green circles If only green circles are present the piece can be accepted while if red circles are present the piece have to be rejected a Fig 24 An example of black dots on a particular 8 References Feigenbaum A V 1986 Total quality Control McGraw Hill 0 07 020353 9 Singapore Gonzales R C R E Woods 1992 Digital image processing Addison Wesley 0 201 50803 6 United States of America Juran J M F M Gryna 1988 eliran s quality control handbook fourth edition McGraw Hill 0 07 033176 6 United States of America Montgomery D
26. lmost immediately If the operator needs to control a production batch it is also possible to acquire several images consecutively and then post process all of them exporting global results in an excel file All this operations are possible thank to the background software that answer to the operator input In fact when the operator select the kind of control the software load all the parameters necessary to the analysis For each application the software load all parameters www intechopen com Digital Image Processing for Quality Control on Injection Molding Products 557 for camera zoom and lights that have been stored earlier and all the information about the analysis like calibration templates for image alignment and so on USER INTERFACE BACKGROUND SOFTWARE DATA STORAGE IMAGE ACQUISITION PARAMETER SETTINGS IMAGE PROCESSING b Fig 1 a Hardware architecture b Software architecture With regard to camera settings the software controls all parameters like time exposure brightness contrast and so on so when the best mix of parameters had been determined for a given application for the analysis is enough to call it back The same happens for the zoom control in fact the software loads the position stored for the selected application and commands to the zoom driver to move in that position This is useful because if necessary the system can acquire images of a large FOV to
27. n example of colour intensity determination of plastic parts for aesthetic goods will be presented This application has the aim to solve the problem of pieces which could appear too dark or too light with respect to a reference one and also to identify defects like undesired striation or black point in the pieces depending on mixing condition of virgin polymer and masterbatch pigment Use of pixel intensity and histograms have been adopted in the development of these procedures 2 System description In this section the hardware and software architecture of the system will be described The hardware is composed by a camera a telecentric zoom lens two lights a support for manual handling of pieces and a PC The camera is a monochromatic camera with a CCD sensor model AVT Stingray F201B The sensor size is 1 1 8 and its resolution is 1624 x 1234 pixel with a pixel dimension of 4 4 pm a colour depth of 8 bit 256 grey levels and a maximum frame rate of 14 fps at full resolution The choice of a CCD sensor increases image quality reducing noise in the acquisition phase and the high resolution about 2 MPixel leads to an acceptable spatial resolution in all the fields of view here adopted The optics of the system is constituted by a telecentric zoom lens model Navitar 12X telecentric zoom the adopted lens is telecentric because it permits to eliminate the perspective error which is a very useful property if one desires to carr
28. nce of time amount Fig 19 www intechopen com 574 Applications and Experiences of Quality Control Time needed for each operation is reported in fig 19 a the values are referred to 8 pieces control Each bar represents the amount of time needed to measure the relative feature It s clear from the picture that the longest control is burr determination with over 5 minutes Then functionality test requires about 3 minutes and all the other feature requires about 1 minutes The amount of time needed is reported in fig 19 c dark blue and is over 13 minutes In automatic control instead the procedure used is different the particular is positioned first in a vertical position in the FOV of the system figure 17 left and the software determines feature D1 D2 D3 and screw termination Then the particular is moved in an horizontal position fig 17 centre and the software determines feature D4 and burr entity The time for change particular position has also been considered Finally functionality tests have been executed as in the previous case Now the longest control is functionality test while horizontal and vertical position requires respectively 1 minute and 1 minute and a half Change position requires about 40 seconds The amount of time needed for the control is about 6 minutes fig 19 c light blue bar The big time reduction near 54 is due to different factors First of all more than a feature can be measured at the sam
29. necessary to locate any feature in the image For example locate the position of an hole respect to the reference system or a reference point given by intersection of two edges figure 10 b shows the process The measurement can be composed as many time as desired including all the feature that can be identified in the FOV www intechopen com Digital Image Processing for Quality Control on Injection Molding Products 565 a Ribbing distance measurement b Hole position determination Fig 10 Process of composed feature measurement 4 Area measurements Digital images are suitable also to measure 2D features such as areas In this section two cases of study will be illustrated In the first example the problem is represented by a burr caused by non perfect adherence between mould parts the aim is to identify and measure the entity of the burr fig 11 a The second problem is again about burr occurrence on the edge as marked with a blue line in fig 11 b but the aim is to measure the area free from burr available for fluid flow The pictures also report the real image of burrs in the parts In the first case burr covers only a part of the edge while in the second example it extends all along the edge circumference b Fig 11 Burrs identification and quantification on two different parts To quantify the entity of the burrs instruments like contour projector are used The only information available from the analysis is the
30. ng for Quality Control on Injection Molding Products 571 The capacity ratio Cp and the performance index Cpx can be evaluated For C evaluation the following equation has been used J M Juran s 1994 o USL LSL 6 60 in which ULS Upper specification Limit LSL Lower specification Limit 60 Process capacity under statistical control The expression used for C instead is J M Juran s 1994 Cor min C Crow 7 where _ USL u i 30 C u LSL low 36 9 As example table 2 reports process performances In this case study the mould has 8 cavity and each cavity has been measured 12 times Totally 96 pieces have been measured Observing C results it is noted that all measured features have an index superior to 1 so the process capability is adequate to meet established tolerance range Moreover the process has high repeatability and feature D1 D2 and D4 have a C that ensure that no pieces can exceed tolerance range in fact Cp is bigger than 1 63 that corresponds to 1 part per million Only feature D3 because of a narrow tolerance range have a C a bit lower alerting that a certain dispersion in production is occurring A further evaluation of results is possible by observing Cpx in fact Cp values are lower evidencing that the process is not centred with respect to the tolerance range Worst situation is yet in feature D3 where Cpk is lower than 1 it means that a considerable percentag
31. ram instead the distribution function is NI Vision Concept Manual 2005 H oiio k ee n 12 where Hcumulaivelk is the number of pixels that are less than or equal to k The cumulated probability function is NI Vision Concept Manual 2005 n Poumulative f z Dro 13 n where Pcumulaive k is the probability that a pixel is less than or equal to k For the application illustrated below only linear histograms will be used 7 1 Colour intensity determination First application has the aim to determine colour intensity of aesthetic goods The problem is due to the production process in fact for this kind of products only a mould is used and different colours of the same particular are obtained mixing masterbatch colour to the virgin polymer and also recycled material like sprue and cold runners If the mixing process is not constant and regular different percentage of colour can be mixed and the particular can be darker or lighter of the master Figure 20 a show the PP cap here examined as example To determine the intensity of the colour the software acquires an image from which extracts a portion figure 20 b shows the acquired image and the extracted area in the yellow rectangle On the extracted image it calculates the histogram and the mean value of gray level intensity of the pixels From figure 21 it is evident that the mean value of the histogram of the dark particular grey level 41 54 is lower than mean value of his
32. rameters are stored for each application 2 2 Image alignment The other operation computed by the software before each analysis is the image alignment This operation simplifies a lot the pieces positioning and make the analysis much easier and faster for the operator In fact the pieces have to be positioned manually in the FOV so it is very difficult to put them always in the same position to permit the software to find features www intechopen com Digital Image Processing for Quality Control on Injection Molding Products 559 for measurement In order to align every image with the master image that was used to develop the specific analysis tool it could be even possible to put a given piece in any position within the field of view and let then the software to rotate and translate the image to match the reference or master image thus to detect features and dimensions However for the sake of accuracy and repeatability the positioning is aided by centring pin and support that permit to place the objects in positions that are similar to the master one used for calibration and parameters setting The alignment procedure is based on pattern or template matching and uses a cross correlation algorithm So first is necessary to define a template fig 4b that the software consider as a feature to find in the image From this template the software extracts pixels that characterize the template shape then it looks for the pixel extracted in the n
33. ss of free area extraction A brief explanation about threshold level must be given smaller is the tolerance range smaller will be the area extracted using a tolerance range of zero only the pixels with a gray level of 255 will be selected and the resulting area will be the smaller as possible In this application the fluctuation is negligible but this must be verified for each case and the most appropriate range has to be selected 5 Shape defects Digital images can also be used not only in measurement applications but also for all controls that involve aesthetic and functionality aspects of moulded parts This kind of controls is generally executed by expert operators that have a depth knowledge of the process and of the part A big experience is required and a proper training for operators is necessary In spite of all the work of qualified operators is not always enough for getting high product quality because often fixed tolerance ranges are not given and the entity of defects can be judged differently from different operators Moreover this kind of defects are frequently not measurable with standard instruments and it is not easy to define a parameter to measure check and certify part quality In this section some examples of shape defects detection by digital images will be showed First will be explained the case of a thread termination damage The part of fig 9 a has a thread in his centre realized by a core that in his term
34. ster image has been extracted and this part is considered as the feature to be searched in the entire image In fig 4 c is showed an image of a piece to align with the master in fact in this image the piece has been moved behind and rotated The software first search the template and determines his rotation angle rotates the image of the same entity to align the new image to the master Then it finds the new position of the template and determines the translation vector from the master moves the image of the fixed quantity and the new image is now aligned to the master Black areas in fig 4 d are the results of translation and rotation of image they are black because these areas have been added by the process and a part of the image has been deleted to keep the same size of the original image A double pattern matching is necessary because of the image reference system is located on the left top of the image and not in the template centre So first pattern matching determines the rotation angle and the translation vector that have to be applied but uses only the first to rotate the image Performing this operation in fact the new image has the same alignment of the master image but the translation vector changes because the rotation is performed respect to the left up corner of the image Second pattern matching determines yet the angle that now is zero and the new translation vector that is used to move the image to the same position of the m
35. t repeatability the results of 100 image measurements on feature D1 as example evidence a standard deviation of 0 0093 mm Better results have been obtained with feature D2 and D4 respectively with standard deviation of 0 0055 mm and 0 0033 mm The best result in repeatability is on D3 feature with standard deviation of 0 0014 mm These values include all possible disturbances that is pixel noise and realignment errors and can be considered as the accuracy of the instrument because of the bias has been deleted choosing acquisition parameter settings that gives the same results as gauge measurements About the process results are reported in table 2 With the exception of feature D3 for which a comparison is not possible it is evident that the average measures present minimum differences This can be expected because the image was calibrated on results of manual measurements Standard deviation also are very similar Moreover once the data has been stored all statistical process evaluations are possible Assuming that the manufacture processes produce pieces whose dimensions or features are statistically scattered according to normal or Gaussian distributions the process capacity can be automatically determined as A V Feigenbaum 1983 2 u u Process Capacity 60 T n where u measurement of int piece u average value of measurements n number of pieces evaluated www intechopen com Digital Image Processi
36. the coordinates of a real reference frame then the mapping can be extended to the entire image Using this method is possible to correct perpendicularity error of camera and scene which is showed in figure 2a This effect is actually rather reduced in the present system as the support has been conceived to provide good mechanical alignment by means of a stiff column that sustains camera in perpendicular position with respect to the scene www intechopen com 558 Applications and Experiences of Quality Control This method of calibration is however useful and must be used to correct also alignment error or rotation error between image axis and real world axis fig 2b It is also possible to define a new reference system in the best point pixel for the studied application for example at the intersection of two edge of the piece analysed 0 0 0 0 eeceocee TEKEK TEKEE TEKET b Rotation error Fig 2 Errors to correct in image calibration About parameter setting the problem is represented by transient materials because the light passes through the material and the dimension measured changes with intensity of illumination So a simple method was used to correct this effect A master of the pieces to analyse was manually measured with calliper and the acquisition parameters particularly brightness contrast ant time exposure of the camera were selected to obtain the same results with digital image measurement All this pa
37. tire particular has been used as template and is www intechopen com Digital Image Processing for Quality Control on Injection Molding Products 569 looked for in inspected image The score of searching procedure is used as parameter to control if the parameter is lower than a threshold fixed to 95 the piece is rejected Ner u a Fig 16 Moulding injection process defects For smaller defects that give no significant changes in score of matching operation it is necessary to adopt methods that analyze only limited parts of the image For example to find air bubble of fig 15 c the software search and identify the edge of the cylinder then transforms points in a line with a linear regression and determines the residual of the operation as Labview 8 2 User Manual 2005 EN A JE 4 where yi represents the true point identified and y represent the point of the fitted line If this parameter is higher than a fixed value about 0 25 mm corresponding to 15 pixel it means that some discontinuity or irregularity is observed on the boundary surface it doesn t matter if outwards or inwards and part is rejected The threshold value has been fixed analysing a sample of 100 good components and taking the bigger value founded The same method has been used for the air bubble on the ribbing but in this case the edges identified and inspected are two the same that can present the defects 6 System performances In this se
38. togram of bright one grey level 64 21 So it is possible to set a range of median value in which the extracted image have to be comprised In this example six ranges have been determined figure 22 from a to f to www intechopen com 5 6 Applications and Experiences of Quality Control separate different intensity colours For each group the median grey level intensity is reported below Consider that six groups represent a resolution approximately three times bigger than human control where only two groups are usually determined and with much more objectivity a The particular examined b The area extracted Fig 20 The particular examined Finally it must be noted that it is possible to deal with coloured parts by means of a monochromatic camera only if one is interested in measuring the colour intensity obviously no considerations can be done on colour tonality So the system is able to detect problems of incorrect mixing as showed in next section or insufficient excessive masterbatch in the mix but it cannot guarantee that the masterbatch used is correct a A light particular b A dark particular m DARK COMPONENT 30 35 40 60 65 70 75 80 c Histograms on comparision E LIGTH COMPONENT Fig 21 Analysis of a light a and a dark b component by histogram c www intechopen com Digital Image Processing for Quality Control on Injection Molding Products 577 a 35 45 b 45 50 c 50 55 d 5
39. values between the edge and the background 2 Edge length defines the maximum distance in which the edge strength has to be verified 3 Edge polarity defines if the greyscale intensity across the edge is rising increase or falling decrease 4 Edge position define x and y location of an edge in the image In picture 7 a is represented a part of an edge black rectangle first the software requires the input of a user defined rectangle green in fig 7 a that it fixes as the ROI in which to look for the edge Then the software divides the rectangle using lines parallel to a rectangle edge in the number specified by the user red arrows in fig 7 a and analyses greyscale value of each pixel line defined moving from the start to the end of the arrow if a rising edge is expected vice versa if a falling edge is expected Now It defines a steepness parameter that represents the region the number of pixels in which the edge strength is expected to verify Then the software averages the pixel value of determinate number width of filter of pixel before and after the point considered The edge strength is computed as the difference between averaged value before and after edge steepness When it finds an edge strength higher than expected edge strength it stores the point and continues with the analysis until the maximum edge strength is reached Now the point found in this way is tagged as edge start and the steepness value is added to find th
40. y out accurate measurements The zoom system permits to have different field of view FOV so the FOV can vary from a minimum of 4 1 mm to a maximum of 49 7 mm The zoom is moved by a stepper driving and the software can communicate and move it automatically to the desired position The utility of this function will be clear later when the start up procedure is explained The depth of focus varies with the FOV ranging from 1 3 mm to 38 8 mm the mentioned camera and lens features bring to a maximum resolution of the system of 0 006 mm FOV 4 1 mm and a minimum resolution of 0 033 mm FOV 49 7 mm To lights the scene a back light and a front light were adopted Both have red light to minimize external noise and to reduce chromatic aberration Moreover they have a coaxial illumination that illumines surface perpendicularly and increases contrast in the image highlighting edge and shapes and improving the general quality of the image In figure 1 a the complete system is showed All the components are managed by a PC that use a specific software developed in LabView Its architecture is reported in figure 1 b It has a user interface that guides step by step the operator through the procedures for the different controls So the operator is called only to choose the application to use then to load pieces in the work area and to shot photos Every time he shots image is acquired and processed and stored if necessary so the result is given a
Download Pdf Manuals
Related Search