工业视觉检查系统中模式识别的研究
|
摘要
本文从经济实用的角度出发，设计了一套工业视觉检查系统。工业视觉大致可分为用于检查、定位和装配三类，本系统属于检查类，是对工业零件进行分类、识别的自动检查装置。在工业生产中，工业环境和照明条件的确定性使得快速准确地判别分类出工业零件的想法成为可能，本系统就是在这种想法下应运而生的。由于在汽车零件生产线上分类产品并进而检查产品合格与否的场合中，要求整个过程在生产线上自动完成，不允许脱线做人工检查，这一检查工作可由本项目研制的工业视觉检查系统自动完成，使得生产更为高效连续。本系统采用图像传感器摄取原始信息，避免了与对象的直接接触和人工检查，所以更符合工业生产在线检查的要求。而且在编制软件时可根据工业零件的形状设计简化的步骤，可达到高效快速的目的。在本课题中，主要进行了对圆形和方形零件分类识别系统的研究和设计，针对的对象是工业零件的强度图像，采用人机对话方式来对零件进行识别。在系统软件设计部分中，首先是对所选零件进行模式识别，包括图像预处理、特征提取和分类器设计三个阶段，其中在图像预处理阶段本系统主要做的工作有：点运算、图像增强、正交变换、边缘提取和边缘增强、轮廓跟踪等。由于视觉系统本身就是一个神经系统，故本文所设计的分类器采用BP神经网络，其具有一些传统技术所没有的优点。但随着应用的广泛，BP网络存在的问题也日益显现出来，主要有：易形成局部极小而得不到整体最优、训练易陷入瘫痪、收敛速度很慢等缺点。本系统针对BP算法的局限性，给出了一种优化的BP算法，采用经过大量实验总结出的经验公式来确定隐层神经元的个数，并选取了一种新的误差平方和函数，该函数的特点是对一些可能的异常点的误差权值设计的较小，从而降低了异常值误差带来的影响，便于模拟出真实的函数关系。采用有导师学习方式，根据样本对在训练过程中的情况，系统地调整其参数。将上述算法应用于对工业零件的识别当中，相对于传统的神经网络可缩短识别时间2.8秒，而且识别正确率可达到81％以上。在对所选零件识别分类之后，本课题又设计了一套零件缺陷检测子系统，缺陷检测正确率可达到91％以上。实验结果表明，若将该系统应用于工业生产线上，可有效地提高生产率和保证产品的质量。
From the view of economy and utility, In this paper, a suit of examination system of industrial vision is designed. Industrial vision can be divided into three classes: inspecting, orientation and setting-up. This system belongs to the class of inspecting, this system is a set of automatic inspecting devices, which is used to classify and recognize industrial parts. During the course of industrial manufacture, the determinacy of the industrial environment and illuminating conditions makes the ideas possible that industrial parts can be distinguished and classified rapidly and accurately. The system introduced below is the result of this idea. In automobile parts product line, if we want to classify products and check them whether they are up to grade or not, all process is required to be finished automatically in the process line and is not permitted to be checked manually on the off-line. The examination system of industrial vision can finished this work automatically and make production faster and more efficient. This system gets original information by image sensor, which avoids direct contact with objects being processed and manual inspection. So this system can meet the requirement that industrial parts should be inspected on-line. At the same time, when software is programmed, simple steps can be designed according to the shape of industrial parts to attain the high-speed and high-efficient purpose. In this task, we presented and designed a system that can classify and recognize circular and quadrate parts and the intension image of industrial parts is the object of the system, human-computer communication mode is adopted to identify the parts. In the part of software design, pattern recognition of selected parts is firstly carried out, it includes three phases: image pretreatment, feature extraction, systematizer design. In the phase of image pretreatment, the main jobs of this system includes dot operation, image swell, positive chiasma transform, edge extraction and edge swell, outline track, etc. Because the visual system itself is a neural system, systematizer designed in the paper adopts BP neural network to accomplish computer image identification, the system has some advantages over the traditional one, but with the extensive application of BP neural network, the problems existing in BP
neural network come forth increasingly. There are some main problems, such as easily forming local extraordinary smallness and not having integral extraordinary excellence, training easily into paralysis, quite slow convergence speed, etc. Aiming at overcoming the limitation of the BP algorithm, this paper brings forward a sort of improved BP algorithm. A experienced equation which is summarized by many experiments is used to determine the number of mesosphere nerve cell and a sort of new square-sum function of errors is adopted. Its characteristic is that weight errors of possible exceptional point is less. Accordingly, the effect of errors of possible exceptional point is reduced, which make actual function relation simulation easier. With having tutor study style, this system adopts a sort of new error function to adjust its parameter systemically according to the status of sample in the course of training. When the algorithm is applied in the identification of industrial parts, comparison with the traditional BP neural network the recognition time will be shortened 2.8 second, and the recognition accuracy can reach more than 81%. After the classification and recognition of selected parts, a disfigurement inspection subsystem is also designed by this task. The accuracy of disfigurement inspection can reach more than 91%. The experiment result indicates, if applying this system to industrial production line can improve the productivity and guarantee the quality of products.
From the view of economy and utility, In this paper, a suit of examination system of industrial vision is designed. Industrial vision can be divided into three classes: inspecting, orientation and setting-up. This system belongs to the class of inspecting, this system is a set of automatic inspecting devices, which is used to classify and recognize industrial parts. During the course of industrial manufacture, the determinacy of the industrial environment and illuminating conditions makes the ideas possible that industrial parts can be distinguished and classified rapidly and accurately. The system introduced below is the result of this idea. In automobile parts product line, if we want to classify products and check them whether they are up to grade or not, all process is required to be finished automatically in the process line and is not permitted to be checked manually on the off-line. The examination system of industrial vision can finished this work automatically and make production faster and more efficient. This system gets original information by image sensor, which avoids direct contact with objects being processed and manual inspection. So this system can meet the requirement that industrial parts should be inspected on-line. At the same time, when software is programmed, simple steps can be designed according to the shape of industrial parts to attain the high-speed and high-efficient purpose. In this task, we presented and designed a system that can classify and recognize circular and quadrate parts and the intension image of industrial parts is the object of the system, human-computer communication mode is adopted to identify the parts. In the part of software design, pattern recognition of selected parts is firstly carried out, it includes three phases: image pretreatment, feature extraction, systematizer design. In the phase of image pretreatment, the main jobs of this system includes dot operation, image swell, positive chiasma transform, edge extraction and edge swell, outline track, etc. Because the visual system itself is a neural system, systematizer designed in the paper adopts BP neural network to accomplish computer image identification, the system has some advantages over the traditional one, but with the extensive application of BP neural network, the problems existing in BP
neural network come forth increasingly. There are some main problems, such as easily forming local extraordinary smallness and not having integral extraordinary excellence, training easily into paralysis, quite slow convergence speed, etc. Aiming at overcoming the limitation of the BP algorithm, this paper brings forward a sort of improved BP algorithm. A experienced equation which is summarized by many experiments is used to determine the number of mesosphere nerve cell and a sort of new square-sum function of errors is adopted. Its characteristic is that weight errors of possible exceptional point is less. Accordingly, the effect of errors of possible exceptional point is reduced, which make actual function relation simulation easier. With having tutor study style, this system adopts a sort of new error function to adjust its parameter systemically according to the status of sample in the course of training. When the algorithm is applied in the identification of industrial parts, comparison with the traditional BP neural network the recognition time will be shortened 2.8 second, and the recognition accuracy can reach more than 81%. After the classification and recognition of selected parts, a disfigurement inspection subsystem is also designed by this task. The accuracy of disfigurement inspection can reach more than 91%. The experiment result indicates, if applying this system to industrial production line can improve the productivity and guarantee the quality of products.
引文
1 Marr D著，姚国正等译．视觉计算理论．科学出版社，1988
2 钟玉琢，乔秉新，李树青．机器人视觉技术．国防工业出版社，1994
3 吴立德．计算机视觉．复旦大学出版社，1993
4 郑南宁．计算机视觉与模式识别．国防工业出版社，1998
5 马颂德，张正友．计算机视觉——计算理论与算法基础．科学出版社，1998
6 边肇祺，张学工．第二版．模式识别．清华大学出版社，2000
7 王积分，张新荣．计算机图像识别．中国铁道出版社，1988
8 陈廷标，夏良正．数字图像处理．人民邮电出版社，1990
9 章毓晋．图像处理和分析(上册)．清华大学出版社，1998
10 [苏]雅罗斯拉夫斯基著，施有秋等译．数字图像处理．电子工业出版社，1993
11 [日]谷口庆治，朱虹等译．数字图像处理．科学出版社，2002
12 Rangarai. M. Rangayyan, Mihai Giu, Farshad Faghih. Adaptive-neighbo rhood filtering of images corrupted by signal-dependent noise. App lied Optics, 1998, (20): 4477～4486
13 佘心平，朱立．一种具有抗噪声干扰的图像边缘提取算法的研究．电子应用技术，1999，(1)：9～10
14 Debendra Nath Das, Sourangshu Mukhopadhyay. Image edge detection and enhancement by an inversion operation. Applied Optics, 1998, (35): 8254～8257
15 Dongqing Cao, Partha P. Banerjeeand Ting-Chung Poon. Image edge en hancement with two cascaded acousto-optic cells with contraprogra pagating sound. Applied Optics, 1998, (14): 3007～3014
16 薛东辉．一种综合的有噪图像边缘提取方法．数据采集与处理，1995，(12)：78～60
17 Amir Shemer, David Mendlovic, Gal Shabtay, Pasacala Garcia-Martin ez, and Javier Garcia. Modified morphological correlation based on bit-map representations. Applied Optics, 1998, (5): 781～787
18 Y. Xu, E. M,-K. lai. Restoration of images contaminated by mixed Gaus sian and impulse noise using a recursive minimum-maximum method. IEEE Proc. Vis. Image Signal Process, 1998, (4): 264～270
19 Pal. N. R, Pal S K. A Review on Image Segmentation Techniques. Patt
ern Recognition, 1993, 26 (9): 1277～1294
20 Yen J C, Chang F J, ChangS. A New Criterion for automatic Multile vel Thresholding. IEEE Trans on Image Processing, 1995, 4 (3): 370～377
21 Sahoo P K, Soltani S, Wang A K C, Chen Y C. A Survey of Thresholding Techniques. Computer Vision. Graphics and Image Processing, 1988, (41): 233～260
22 Pikaz A, Averbuch A, Digital Image Thresholding Based on Topological Stable State. Pattern Recognition, 1996, 29(5): 829～843
23 Papamarkos N, Gatos B. A New Approach for Multilevel Threshold Selection. CVGIP: Graphic Models and Image Processing, 1994, (56): 357～370
24 Corneloup G, Moysan J, Maynin I E. BSCAN Image Segmentation by Thresholding Using Cooccurrence Matrix Analysis, Pattern Recognition, 1996, 29(2): 281～296
25 Li L, Gong J, Chen W. Grey-Level Image Thresholding Based on Fisher Linear Projection of Two-Dimensional Histogram. Pattern Recogn ition, 1997, 30(5): 743～750
26 Brink A D. Thresholding of Digital Images Using Two-Dimensional Entropies. Pattern Recognition, 1992, 25(8): 803～808
27 张远鹏等．计算机图像处理技术基础．北京大学出版社，1996
28 Haralick. R.M. Digital Step Edges from Zero Crossing of Second Directional Derivatives. IEEE Trans on Pattern Analysis and Machine Intelligence, 1984, 6(1): 58～68
29 Clark J J. Authenticating Edges Produced by Zero-Crossing Algorithm. IEEE Trans on Pattern Analysis and Machine Intelligence, 1989, 11(1): 43～57
30 Wilkinson M H F. Optimizing Edge Detectors for Robust Automatic T hreshold Selection: Coping with Edge Curvature and Noise. Graphic Models and Image Processing, 1998, 60(4): 385～401
31 日本电视协会．图像与噪声．人民邮电出版社，1984
32 Babaud J, Andrew P, Bardin W M, Duda R.O. Uniqueness of the Gausian Kernel for Scale-Space Filtering. IEEE Trans on Pattern Analysis and
Machine Intelligence, 1986, 8(1): 26～33
33 Elder J H, Zucher S W. Local Scale Control for Edge Detection and Blur Estimation. IEEE Trans on Pattern Analysis and Machine Intelligence, 1998, 20(7): 699～716
34 余松煜，周源华，吴时光．数字图像处理．电子工业出版社，1989
35 夏德深，傅德胜．现代图像处理技术．东南大学出版社，1997
36 Gatte F, Lions P L, Morel J M, Coll T. Image Selective Smoothing and Edge Detection by Nonlinear Diffusion I. SIAM. J Numer Anal, 1992, 29(1): 182～193
37 Gatte F, Lions P L, Morel J M, Coll T. Image Selective Smoothing and Edge Detection by Nonlinear Diffusion Ⅱ. SIAM. J Numer Anal, 1992, 29(3): 845～866
38 Nitzberg M, Shiota T. Nonliear Image Filtering with Edge and Corner Enhancement. IEEE Trans on Pattern Analysis and Machine Intelligence, 1992, 14(8): 826～833
39 Legault R, Suen C Y. Optimal Local Weighted Averaging Methods in Contour Smoothing. IEEE Trans on Pattern Analysis and Machine Intelligence, 1997, 19(9): 801～817
40 Pavlidis T, Liow Y T. Integrating Region Growing and Edge Detection. IEEE Trans on Pattern Analysis and Machine Intelligence, 1990, 12(3): 225～233
41 焦李成．神经网络系统理论．西安电子科技大学出版社，1999
42 何斌，马天予等．Visual C++数字图像处理．人民邮电出版社，2001
43 周孝宽，曹晓光等．实用微机图像处理．北京航天航空大学出版社，1994
44 Kenneth．R．Castleman，朱志刚等译．数字图像处理．电子工业出版社，1998
45 沈清，汤霖．模式识别导论．国防科技大学出版社，1991
46 易伟文．图形识别的神经网络方法．微机发展，1996，(20)：28～31
47 孙毅辉．基于神经网络的图像识别系统．应用技术，2001，(3)：13～15
48 徐华，刘红琳．BP神经网络学习参数的自调整技术及其应用．网络天地，1997，(3)：40～41
49 彭天好，范龙振．BP神经网络的一种稳健改进方法．计算机应用研究，1996，(6)：29～30．
50 王耀南．一种神经网络的快速学习算法及其在图像边缘识别中的应用．计算机研究与发展，1997，34(5)：377～381
51 柳文，冯建农等．一种改进型快速BP训练算法．计算机工程与科学，1998，20(11)：17～20
52 严卫，朱兆达等．BP网络进化及其在雷达目标识别中的应用．数据采集与处理，1999，(9)：273～277
53 向国全，董道珍．BP模型中的激励函数和改进的网络训练法．计算机研究与发展，1997，34(2)：113～117
54 江虹，曾立波等．优化的BP神经网络分类器的设计与实现．计算机工程与应用，2001，(5)：122～125
2 钟玉琢，乔秉新，李树青．机器人视觉技术．国防工业出版社，1994
3 吴立德．计算机视觉．复旦大学出版社，1993
4 郑南宁．计算机视觉与模式识别．国防工业出版社，1998
5 马颂德，张正友．计算机视觉——计算理论与算法基础．科学出版社，1998
6 边肇祺，张学工．第二版．模式识别．清华大学出版社，2000
7 王积分，张新荣．计算机图像识别．中国铁道出版社，1988
8 陈廷标，夏良正．数字图像处理．人民邮电出版社，1990
9 章毓晋．图像处理和分析(上册)．清华大学出版社，1998
10 [苏]雅罗斯拉夫斯基著，施有秋等译．数字图像处理．电子工业出版社，1993
11 [日]谷口庆治，朱虹等译．数字图像处理．科学出版社，2002
12 Rangarai. M. Rangayyan, Mihai Giu, Farshad Faghih. Adaptive-neighbo rhood filtering of images corrupted by signal-dependent noise. App lied Optics, 1998, (20): 4477～4486
13 佘心平，朱立．一种具有抗噪声干扰的图像边缘提取算法的研究．电子应用技术，1999，(1)：9～10
14 Debendra Nath Das, Sourangshu Mukhopadhyay. Image edge detection and enhancement by an inversion operation. Applied Optics, 1998, (35): 8254～8257
15 Dongqing Cao, Partha P. Banerjeeand Ting-Chung Poon. Image edge en hancement with two cascaded acousto-optic cells with contraprogra pagating sound. Applied Optics, 1998, (14): 3007～3014
16 薛东辉．一种综合的有噪图像边缘提取方法．数据采集与处理，1995，(12)：78～60
17 Amir Shemer, David Mendlovic, Gal Shabtay, Pasacala Garcia-Martin ez, and Javier Garcia. Modified morphological correlation based on bit-map representations. Applied Optics, 1998, (5): 781～787
18 Y. Xu, E. M,-K. lai. Restoration of images contaminated by mixed Gaus sian and impulse noise using a recursive minimum-maximum method. IEEE Proc. Vis. Image Signal Process, 1998, (4): 264～270
19 Pal. N. R, Pal S K. A Review on Image Segmentation Techniques. Patt
ern Recognition, 1993, 26 (9): 1277～1294
20 Yen J C, Chang F J, ChangS. A New Criterion for automatic Multile vel Thresholding. IEEE Trans on Image Processing, 1995, 4 (3): 370～377
21 Sahoo P K, Soltani S, Wang A K C, Chen Y C. A Survey of Thresholding Techniques. Computer Vision. Graphics and Image Processing, 1988, (41): 233～260
22 Pikaz A, Averbuch A, Digital Image Thresholding Based on Topological Stable State. Pattern Recognition, 1996, 29(5): 829～843
23 Papamarkos N, Gatos B. A New Approach for Multilevel Threshold Selection. CVGIP: Graphic Models and Image Processing, 1994, (56): 357～370
24 Corneloup G, Moysan J, Maynin I E. BSCAN Image Segmentation by Thresholding Using Cooccurrence Matrix Analysis, Pattern Recognition, 1996, 29(2): 281～296
25 Li L, Gong J, Chen W. Grey-Level Image Thresholding Based on Fisher Linear Projection of Two-Dimensional Histogram. Pattern Recogn ition, 1997, 30(5): 743～750
26 Brink A D. Thresholding of Digital Images Using Two-Dimensional Entropies. Pattern Recognition, 1992, 25(8): 803～808
27 张远鹏等．计算机图像处理技术基础．北京大学出版社，1996
28 Haralick. R.M. Digital Step Edges from Zero Crossing of Second Directional Derivatives. IEEE Trans on Pattern Analysis and Machine Intelligence, 1984, 6(1): 58～68
29 Clark J J. Authenticating Edges Produced by Zero-Crossing Algorithm. IEEE Trans on Pattern Analysis and Machine Intelligence, 1989, 11(1): 43～57
30 Wilkinson M H F. Optimizing Edge Detectors for Robust Automatic T hreshold Selection: Coping with Edge Curvature and Noise. Graphic Models and Image Processing, 1998, 60(4): 385～401
31 日本电视协会．图像与噪声．人民邮电出版社，1984
32 Babaud J, Andrew P, Bardin W M, Duda R.O. Uniqueness of the Gausian Kernel for Scale-Space Filtering. IEEE Trans on Pattern Analysis and
Machine Intelligence, 1986, 8(1): 26～33
33 Elder J H, Zucher S W. Local Scale Control for Edge Detection and Blur Estimation. IEEE Trans on Pattern Analysis and Machine Intelligence, 1998, 20(7): 699～716
34 余松煜，周源华，吴时光．数字图像处理．电子工业出版社，1989
35 夏德深，傅德胜．现代图像处理技术．东南大学出版社，1997
36 Gatte F, Lions P L, Morel J M, Coll T. Image Selective Smoothing and Edge Detection by Nonlinear Diffusion I. SIAM. J Numer Anal, 1992, 29(1): 182～193
37 Gatte F, Lions P L, Morel J M, Coll T. Image Selective Smoothing and Edge Detection by Nonlinear Diffusion Ⅱ. SIAM. J Numer Anal, 1992, 29(3): 845～866
38 Nitzberg M, Shiota T. Nonliear Image Filtering with Edge and Corner Enhancement. IEEE Trans on Pattern Analysis and Machine Intelligence, 1992, 14(8): 826～833
39 Legault R, Suen C Y. Optimal Local Weighted Averaging Methods in Contour Smoothing. IEEE Trans on Pattern Analysis and Machine Intelligence, 1997, 19(9): 801～817
40 Pavlidis T, Liow Y T. Integrating Region Growing and Edge Detection. IEEE Trans on Pattern Analysis and Machine Intelligence, 1990, 12(3): 225～233
41 焦李成．神经网络系统理论．西安电子科技大学出版社，1999
42 何斌，马天予等．Visual C++数字图像处理．人民邮电出版社，2001
43 周孝宽，曹晓光等．实用微机图像处理．北京航天航空大学出版社，1994
44 Kenneth．R．Castleman，朱志刚等译．数字图像处理．电子工业出版社，1998
45 沈清，汤霖．模式识别导论．国防科技大学出版社，1991
46 易伟文．图形识别的神经网络方法．微机发展，1996，(20)：28～31
47 孙毅辉．基于神经网络的图像识别系统．应用技术，2001，(3)：13～15
48 徐华，刘红琳．BP神经网络学习参数的自调整技术及其应用．网络天地，1997，(3)：40～41
49 彭天好，范龙振．BP神经网络的一种稳健改进方法．计算机应用研究，1996，(6)：29～30．
50 王耀南．一种神经网络的快速学习算法及其在图像边缘识别中的应用．计算机研究与发展，1997，34(5)：377～381
51 柳文，冯建农等．一种改进型快速BP训练算法．计算机工程与科学，1998，20(11)：17～20
52 严卫，朱兆达等．BP网络进化及其在雷达目标识别中的应用．数据采集与处理，1999，(9)：273～277
53 向国全，董道珍．BP模型中的激励函数和改进的网络训练法．计算机研究与发展，1997，34(2)：113～117
54 江虹，曾立波等．优化的BP神经网络分类器的设计与实现．计算机工程与应用，2001，(5)：122～125