高精度视觉检测系统的研究
摘要
机器视觉检测是基于机器视觉技术、光学测量原理形成的一种新型检测技术,被广泛应用于几何量的尺寸测量、航空遥感测量、高精度定位、精密复杂零件的微尺寸测量和零件外观检测等与图像有关的技术领域中。如何提高检测精度成为该领域的重要研究课题。
本文论述了视觉检测系统的检测原理、组成,分析了影响系统精度的主要因素,并提出了消除或者减少误差的有效方法。分析了现有的亚像素检测技术,提出了一种亚像素边缘检测方法——基于扩展的sobel边缘的最小二乘曲线拟合法,大大提高了获取边缘的精度。针对图像中特征线,提出了一种改进的hough变换方法——基于边缘区域的斜率统计法来提取直线,采用了三点定位圆法来提取圆及其圆心,提高了计算速度、定位精度。构建了插线卡视觉检测系统,运用提出的算法,提高了系统的检测精度。
Modern science and technology tends to develop rapidly into small, ultra-precision fields.It develops from mm and micron level to nano technology. With Micro、Nano Technology rising , it has already caused the evolutional change ,such as manufacture material, biology, communication, national defence,bringing about huge influence on the living quality of human and society. With the continuous development of micro-technology, the characteristic of small-scale micro-mechanical has become an advanced technique of people knowing and reforming objective world from the tiny view angle. With enhancing of the industry manufacture technology and the machine process, people put forward of higher requests to the product’s detecting approaches, detecting speed and detecting precision. But exiting inspection means can’t give attention to either speed or inspection. But inspection’s precision and efficiency decided the development of level of manufacturing and technology in some extent. So we must find a new inspection technique to resolve this problem. As a result, it brings a new technique——the technique of high accuracy of vision inspection.
High accuracy of Machine vision inspection is a new technique based on modern optics in measurement area, and it is a modern test technique which includes optical electronics, computer iconography, information processing and computer vision etc. it regards picture as means and carrier of inspection and passing information and pick up useful information from picture and acquire all kind of parameter needly by deal with picture , vision examination technique’s have many advantage ,such as no touch, all vision and high speed, it also has some advantage ,such as on line examination real time analyze , real time control. It is widely used in military industry, inspection of product ,high accuracy measurement , medicine and so on.
Since the application of visual system has used from 1980s,it is ued very wide.The computer vision examination has obtained the vigorous development, the new concept, the new theory, the new method have unceasingly emerged, and it has been widely used in various fields.But In the domestic, the industry vision system still is in the conceptual phase.compared with overseas,the domestic computer vision examination research and the application level has 20 years disparity at least.So, in further exploring the theory and practice of visual inspection, absorbing advanced technology and academic thought, broadening the use of fields of visual inspection are very important.
Combinging project of industry key scientific research(the project serial number: 2005B100014) of the NingBo city-- the research of small size measurement system based on machine vision ,The article has studied the examination principle and therealization method of the high accuracy vision examination system, and analyzed the primary factor of influencing system’s precision, And proposed the way of increasing the system precision. Finally we taked stip as the examination object, and constructed the high accuracy examination system, and carried on the examination according to the technical requirements of testing components.
The full text of main content as follows:
This article discussed the general pattern of high-precision machine vision inspection system, the system works, given the hardware components of visual detection system,and detailed the selection description of the detection system hardware, stressed the importance of the software and analysis key technologies of influence the the precision of machine vision examination system.
The main task of Visual detection system is image gathering,image processing. Images have been affected with error in the collection, transmission, processing process and so on, and these errors are inevitable. These seerrors will have a serious impact on the system accuracy until us analysis and processe to these seerrors.The papers analyzes the primary factor of influencing system’s precision, and proposed the reduction or elimination of these error methods to improve the measurement accuracy of the system.
Sub-pixel edge detection technology is a new development of image processing technology, and a high precision vision inspection technology. It occupies an important position in the field of vision inspection. Image Feature detection relates the ultimate success or failure of detection. Weather effectively extracting the necessary geometric features is the important of task of vision detection. The paper describes the concept of sub-pixel edge detection and explores the sub-pixel edge detection method which is used widly such as cubic spline interpolation and Curve fitting and so on.A method of Least squares curve-fitting interpolation based on the edge of a sobel method is proposed to improve the accuracy and computational speed of the edge detection. Hough transform is a traditional way which computational speed is lowly and positioning accuracy is not high. This paper also proposed a method of slope statistics based on the edge region to extract a straight line. Proposed a method of three points to extract the round and center of round makes the amount of calculation reduce and also makes detection more accurate. Experiments show that the methods proposed by paper have fast and high accuracy, and improve accuracy of the system effectively.
The paper has constructed vision detection system of the clip, analyzed the detected task of clip and the detection principle, selected the appropriate hardware, and developed software of measureing parameter of clip, carried on examination on-line.it spends two seconds to measurement parameters for the entire testing period. The detection system has a very strong usability to many tasks which has very similar detection task and also has the bigger promoted value.
本文论述了视觉检测系统的检测原理、组成,分析了影响系统精度的主要因素,并提出了消除或者减少误差的有效方法。分析了现有的亚像素检测技术,提出了一种亚像素边缘检测方法——基于扩展的sobel边缘的最小二乘曲线拟合法,大大提高了获取边缘的精度。针对图像中特征线,提出了一种改进的hough变换方法——基于边缘区域的斜率统计法来提取直线,采用了三点定位圆法来提取圆及其圆心,提高了计算速度、定位精度。构建了插线卡视觉检测系统,运用提出的算法,提高了系统的检测精度。
Modern science and technology tends to develop rapidly into small, ultra-precision fields.It develops from mm and micron level to nano technology. With Micro、Nano Technology rising , it has already caused the evolutional change ,such as manufacture material, biology, communication, national defence,bringing about huge influence on the living quality of human and society. With the continuous development of micro-technology, the characteristic of small-scale micro-mechanical has become an advanced technique of people knowing and reforming objective world from the tiny view angle. With enhancing of the industry manufacture technology and the machine process, people put forward of higher requests to the product’s detecting approaches, detecting speed and detecting precision. But exiting inspection means can’t give attention to either speed or inspection. But inspection’s precision and efficiency decided the development of level of manufacturing and technology in some extent. So we must find a new inspection technique to resolve this problem. As a result, it brings a new technique——the technique of high accuracy of vision inspection.
High accuracy of Machine vision inspection is a new technique based on modern optics in measurement area, and it is a modern test technique which includes optical electronics, computer iconography, information processing and computer vision etc. it regards picture as means and carrier of inspection and passing information and pick up useful information from picture and acquire all kind of parameter needly by deal with picture , vision examination technique’s have many advantage ,such as no touch, all vision and high speed, it also has some advantage ,such as on line examination real time analyze , real time control. It is widely used in military industry, inspection of product ,high accuracy measurement , medicine and so on.
Since the application of visual system has used from 1980s,it is ued very wide.The computer vision examination has obtained the vigorous development, the new concept, the new theory, the new method have unceasingly emerged, and it has been widely used in various fields.But In the domestic, the industry vision system still is in the conceptual phase.compared with overseas,the domestic computer vision examination research and the application level has 20 years disparity at least.So, in further exploring the theory and practice of visual inspection, absorbing advanced technology and academic thought, broadening the use of fields of visual inspection are very important.
Combinging project of industry key scientific research(the project serial number: 2005B100014) of the NingBo city-- the research of small size measurement system based on machine vision ,The article has studied the examination principle and therealization method of the high accuracy vision examination system, and analyzed the primary factor of influencing system’s precision, And proposed the way of increasing the system precision. Finally we taked stip as the examination object, and constructed the high accuracy examination system, and carried on the examination according to the technical requirements of testing components.
The full text of main content as follows:
This article discussed the general pattern of high-precision machine vision inspection system, the system works, given the hardware components of visual detection system,and detailed the selection description of the detection system hardware, stressed the importance of the software and analysis key technologies of influence the the precision of machine vision examination system.
The main task of Visual detection system is image gathering,image processing. Images have been affected with error in the collection, transmission, processing process and so on, and these errors are inevitable. These seerrors will have a serious impact on the system accuracy until us analysis and processe to these seerrors.The papers analyzes the primary factor of influencing system’s precision, and proposed the reduction or elimination of these error methods to improve the measurement accuracy of the system.
Sub-pixel edge detection technology is a new development of image processing technology, and a high precision vision inspection technology. It occupies an important position in the field of vision inspection. Image Feature detection relates the ultimate success or failure of detection. Weather effectively extracting the necessary geometric features is the important of task of vision detection. The paper describes the concept of sub-pixel edge detection and explores the sub-pixel edge detection method which is used widly such as cubic spline interpolation and Curve fitting and so on.A method of Least squares curve-fitting interpolation based on the edge of a sobel method is proposed to improve the accuracy and computational speed of the edge detection. Hough transform is a traditional way which computational speed is lowly and positioning accuracy is not high. This paper also proposed a method of slope statistics based on the edge region to extract a straight line. Proposed a method of three points to extract the round and center of round makes the amount of calculation reduce and also makes detection more accurate. Experiments show that the methods proposed by paper have fast and high accuracy, and improve accuracy of the system effectively.
The paper has constructed vision detection system of the clip, analyzed the detected task of clip and the detection principle, selected the appropriate hardware, and developed software of measureing parameter of clip, carried on examination on-line.it spends two seconds to measurement parameters for the entire testing period. The detection system has a very strong usability to many tasks which has very similar detection task and also has the bigger promoted value.
引文
[1] 贾德云. 机器视觉. 科学技术出版社, 2002.
[2] 郑南宁. 计算机视觉与模式识别. 国防工业出版社, 1998.3.
[3] 李大友,刘小伟. 图形图像处理技术. 清华大学出版社, 2004.5.
[4] Franci Lahajnar et al. Machine vision system for inspecting electric plates. Computer in Industry. 2002, 47:113～122.
[5] Perkins, W. A. Model-based inspection system for component boards. Pattern Recognition. 1984, 17: 135～140.
[6] Kap Tuk Chan et al. Identifying IC Type from Its Package using Character Recognition Technique. SPIE Vol. 3185: 50～57.
[7] TIMOTHYS.NEWWAN and ANILK.JAIN. A system for 3d cad-based inspection using range images. Pattern Recognition . 1995, 28(10): 1555～1574.
[8] Sternberg, S.R. Industrial inspection by morphological virtual gauging. Computer Society Workshop on Computer Architectures for Pattern Analysis.1983: 237～247.
[9] Dae-ki kang et al. Applying object modeling technique to automated visual inspection of automotive compressor parts omission. International Journal of Machine tools&manufacture.1999, 39: 1779～1792.
[10] 王庆有. CCD 应用技术[M]. 天津大学出版社,2000.
[11] Kim,Pyunghyun,Rhee,Sehun.Three dimensional inspection of ball gri d array using laser vision system[J].IEEE Transactions on Electronics Packaging Manufacturing,1999,22(2)151 - 155.
[12] C.J.Tay,He.X.,Kang.X.Coplanarity study on ball grid array packagi ng [J] Optics Communications,2001,40(8)1608 - 1612.
[13] 杨丽凤,韩冀皖,李元宗. 面阵高精度测量技术的应用.太原理工大学学报.2001,32(5):455-458.
[14] 左奇,史忠科.基于机器视觉的胶囊完整性检测系统研究.西安交通大学学报.2002,36.
[15] 阮秋琦. 数字图像处理. 电子工业出版社, 2004.
[16] 朱铮涛. 基于计算机视觉图像精密测量的关键技术研究. 华南理工大学博士学位论文.
[17] 林凡,吴孙桃, 郭东辉. CMOS 图像传感器技术及其研究进展. 半导体技术 , 2001 (12).
[18] 王跃科,杨华勇. CCD 图像传感技术的现状与应用前景[J] . 光学仪器,199 6 ,18(5):32 - 36.
[19] 科学 CCD 的过去、现状和未来[J]. 激光与光电子学进展, 1995 , (10):8 - 10.
[20] 王庆有,孙学珠. CCD 应用技术. 天津大学出版社,1993.
[21] Weng J,Cohen P,Herniou M.Camera Calibration with Distortion Models and Accuracy Evaluation.PA-MI,1992,14(10):965~980.
[22] 吴晓波,安文斗. 图像测量系统中的误差分析及提高测量精度的途径. 精密工程, 1997,2.
[23] 吴晓波,钟先信. 高精度图像测量系统. 光学精密工程,1995,3(1):20-33.
[24] 吴晓波. 提高面阵 CCD 尺寸测量分辨率的新途径. 中国科协第二届青年学术年会四川卫星会议论文集,1995 上册:656~659.
[25] 章毓晋. 图像工程(下册). 北京:清华大学出版社, 2000.
[26] 吴晓波,钟先信. 应用多项式插值函数提高面阵 CCD 尺寸测量的分辨力[J].仪器仪表学报,1996,17(2):154-158.
[27] ZOU Ning ,LIU Jian ,ZHOU Man2li. Zernike moment2basededge detection [ J ] . Journal Huazhong University of Sci2ence and Technology( 华中科技大学学报) ,1999 ,27 (5) :122 ,8. (in Chine se).
[28] 郭彦珍, 尹国鑫. 应用概率论提高 CCD 尺寸测量的分辨力[J]. 仪器仪表学报,1988,9(2):149-154.
[29] LI Yan , PANG Jia2xiong. Sub2pixel edge detection basedon spline interpolation of D2 and LOGoperator [J]. JournaHuazhong University of Science and Technology (华中科技大学学报) ,2000 ,28 (3) :77279. (in Chine se).
[30] 杨烜,梁德群. 基于边缘视觉特征描述的自适应多尺度小波边缘检测方法.电子科学学刊,1998,20(5).
[31] 徐涛. 数值计算方法. 吉林科技出版社.1998.
[32] E.P.Lyvers,O.R.Mitchell. Precision edge contrast and orientation estimation[J] IEEETrans.onPAMI,1988,10(6):927~937.
[33] 贺忠海,王宝光,廖怡白,陈林才. 利用曲线拟合方法的亚像素提取算法. 仪器仪表学报,2003年4月.
[34] Xu, E Oja, P Kultanan.A new curve detection method; Randomized Hough transform RHT.Patten Recog Len 11.1990:331-338.
[35] 张健新,段发阶,叶声华. 简便的高精度摄像机标定技术.仪器仪表学报[J],1999,20(2):193-196.
[36] 孙长库,魏嵬,张效栋等. CCD 摄像机参数标定实验设计. 光电子技术与信息[J],2005,18(2):43-46.
[37] 中科院自动化研究所模式识别国家重点实验室资料. 摄像机标定;
[38] 杨年峰,王季军,黄昌华等. 线性变换法中标定对三维重构精度的影响. 清华大学学报(自然科学版)[J], 2000,40(4):24-27.
[39] Tsai R Y. A versatile camera calibration technique for high-accuracy 3D machine vision metrology using off-the-shelf TV cameras and lenses[J]. IEEE Journal of Robotics and Automation, 1987, 3(4):323-343.
[40] 张正友. 计算机视觉——计算理论与算法基础. 科学出版社, 1998.
[41] 吴文琪,孙增圻. 机器视觉中的摄像机定标方法综述. 计算机应用研究[J].2004,2:4-6.
[42] 胡占义,吴福朝. 基于主动视觉摄像机标定方法. 计算机学报[J], 2002, 25(11): 1149-1156.
[43] 吴福朝,胡占义. 摄像机自标定的线性理论与算法. 计算机学报[J],2001, 24(11): 1121-1135.
[44] 吴福朝,胡占义. 线性确定无穷远平面单应矩阵的确定和摄像机标定. 自动化学报[J], 2002, 28(4): 488-496.
[45] 孟晓桥,胡占义. 摄像机自标定方法的研究与进展. 自动化学报[J],2003, 29(1): 110-124.
[46] 夏良正,李久贤. 数字图像处理. 东南大学出版社, 2005.
[47] 陆宗骐. C++ 图像处理编程. 清华大学出版社, 2005.
[48] 李晶皎,朱志良. 模式识别. 电子工业出版社, 2004.
[49] 朱虹. 数字图像处理. 科学出版社, 2005.
[50] Kenneth. R..Castleman . Digital image processing. 清华大学出版社, 2003.
[51] 谷口庆治. 数字图像处理. 科学出版社,2002.
[52] 董长虹等. Matlab 图像处理与应用. 国防工业出版社, 2004.
[53] 成礼智,王红霞等. 小波的理论与应用. 科学出版社, 2004.
[54] 黄爱民,安向京,骆力等. 数字图像处理与分析基础. 中国水利水电出版社, 2005.
[55] 张宏林,蔡锐等. VC++数字图像模式识别技术及工程实践. 人民邮电出版社, 2003.
[56] 龚建伟,熊光明. Visual C++ rbo C 串口通信编程实践. 电子工业出版社2004-10-1.
[57] 吴晓波. 应用多项式插值函数提高面阵 CCD 尺寸测量的分辨力. 《仪器仪表学报》,1996 / 17 / 02 P 154-158.
[58] 郭彦珍. 应用概率论提高 CCD 尺寸测量的分辨力. 《 仪器仪表学报》 , 1988 / 9 / 02 P 149-154.
[59] 邹仲力. 提高 CCD 尺寸测量分辨力的解调测量法[J]. 仪器仪表学报,1986,7(1):38~45.
[60] 韩军伟. 景物边缘提取技术的研究与应用. 西北工业大学硕士学位论文,2001 年 3 月:24～39.
[61] 张艳珍. 微机视觉相关理论及技术研究. 大连理工大学博士学位论文.2001 年 4 月:16～35.
[62] 王富治. 高精度视觉测量系统研究. 电子科技大学硕士学位论文. 2005.6.
[2] 郑南宁. 计算机视觉与模式识别. 国防工业出版社, 1998.3.
[3] 李大友,刘小伟. 图形图像处理技术. 清华大学出版社, 2004.5.
[4] Franci Lahajnar et al. Machine vision system for inspecting electric plates. Computer in Industry. 2002, 47:113～122.
[5] Perkins, W. A. Model-based inspection system for component boards. Pattern Recognition. 1984, 17: 135～140.
[6] Kap Tuk Chan et al. Identifying IC Type from Its Package using Character Recognition Technique. SPIE Vol. 3185: 50～57.
[7] TIMOTHYS.NEWWAN and ANILK.JAIN. A system for 3d cad-based inspection using range images. Pattern Recognition . 1995, 28(10): 1555～1574.
[8] Sternberg, S.R. Industrial inspection by morphological virtual gauging. Computer Society Workshop on Computer Architectures for Pattern Analysis.1983: 237～247.
[9] Dae-ki kang et al. Applying object modeling technique to automated visual inspection of automotive compressor parts omission. International Journal of Machine tools&manufacture.1999, 39: 1779～1792.
[10] 王庆有. CCD 应用技术[M]. 天津大学出版社,2000.
[11] Kim,Pyunghyun,Rhee,Sehun.Three dimensional inspection of ball gri d array using laser vision system[J].IEEE Transactions on Electronics Packaging Manufacturing,1999,22(2)151 - 155.
[12] C.J.Tay,He.X.,Kang.X.Coplanarity study on ball grid array packagi ng [J] Optics Communications,2001,40(8)1608 - 1612.
[13] 杨丽凤,韩冀皖,李元宗. 面阵高精度测量技术的应用.太原理工大学学报.2001,32(5):455-458.
[14] 左奇,史忠科.基于机器视觉的胶囊完整性检测系统研究.西安交通大学学报.2002,36.
[15] 阮秋琦. 数字图像处理. 电子工业出版社, 2004.
[16] 朱铮涛. 基于计算机视觉图像精密测量的关键技术研究. 华南理工大学博士学位论文.
[17] 林凡,吴孙桃, 郭东辉. CMOS 图像传感器技术及其研究进展. 半导体技术 , 2001 (12).
[18] 王跃科,杨华勇. CCD 图像传感技术的现状与应用前景[J] . 光学仪器,199 6 ,18(5):32 - 36.
[19] 科学 CCD 的过去、现状和未来[J]. 激光与光电子学进展, 1995 , (10):8 - 10.
[20] 王庆有,孙学珠. CCD 应用技术. 天津大学出版社,1993.
[21] Weng J,Cohen P,Herniou M.Camera Calibration with Distortion Models and Accuracy Evaluation.PA-MI,1992,14(10):965~980.
[22] 吴晓波,安文斗. 图像测量系统中的误差分析及提高测量精度的途径. 精密工程, 1997,2.
[23] 吴晓波,钟先信. 高精度图像测量系统. 光学精密工程,1995,3(1):20-33.
[24] 吴晓波. 提高面阵 CCD 尺寸测量分辨率的新途径. 中国科协第二届青年学术年会四川卫星会议论文集,1995 上册:656~659.
[25] 章毓晋. 图像工程(下册). 北京:清华大学出版社, 2000.
[26] 吴晓波,钟先信. 应用多项式插值函数提高面阵 CCD 尺寸测量的分辨力[J].仪器仪表学报,1996,17(2):154-158.
[27] ZOU Ning ,LIU Jian ,ZHOU Man2li. Zernike moment2basededge detection [ J ] . Journal Huazhong University of Sci2ence and Technology( 华中科技大学学报) ,1999 ,27 (5) :122 ,8. (in Chine se).
[28] 郭彦珍, 尹国鑫. 应用概率论提高 CCD 尺寸测量的分辨力[J]. 仪器仪表学报,1988,9(2):149-154.
[29] LI Yan , PANG Jia2xiong. Sub2pixel edge detection basedon spline interpolation of D2 and LOGoperator [J]. JournaHuazhong University of Science and Technology (华中科技大学学报) ,2000 ,28 (3) :77279. (in Chine se).
[30] 杨烜,梁德群. 基于边缘视觉特征描述的自适应多尺度小波边缘检测方法.电子科学学刊,1998,20(5).
[31] 徐涛. 数值计算方法. 吉林科技出版社.1998.
[32] E.P.Lyvers,O.R.Mitchell. Precision edge contrast and orientation estimation[J] IEEETrans.onPAMI,1988,10(6):927~937.
[33] 贺忠海,王宝光,廖怡白,陈林才. 利用曲线拟合方法的亚像素提取算法. 仪器仪表学报,2003年4月.
[34] Xu, E Oja, P Kultanan.A new curve detection method; Randomized Hough transform RHT.Patten Recog Len 11.1990:331-338.
[35] 张健新,段发阶,叶声华. 简便的高精度摄像机标定技术.仪器仪表学报[J],1999,20(2):193-196.
[36] 孙长库,魏嵬,张效栋等. CCD 摄像机参数标定实验设计. 光电子技术与信息[J],2005,18(2):43-46.
[37] 中科院自动化研究所模式识别国家重点实验室资料. 摄像机标定;
[38] 杨年峰,王季军,黄昌华等. 线性变换法中标定对三维重构精度的影响. 清华大学学报(自然科学版)[J], 2000,40(4):24-27.
[39] Tsai R Y. A versatile camera calibration technique for high-accuracy 3D machine vision metrology using off-the-shelf TV cameras and lenses[J]. IEEE Journal of Robotics and Automation, 1987, 3(4):323-343.
[40] 张正友. 计算机视觉——计算理论与算法基础. 科学出版社, 1998.
[41] 吴文琪,孙增圻. 机器视觉中的摄像机定标方法综述. 计算机应用研究[J].2004,2:4-6.
[42] 胡占义,吴福朝. 基于主动视觉摄像机标定方法. 计算机学报[J], 2002, 25(11): 1149-1156.
[43] 吴福朝,胡占义. 摄像机自标定的线性理论与算法. 计算机学报[J],2001, 24(11): 1121-1135.
[44] 吴福朝,胡占义. 线性确定无穷远平面单应矩阵的确定和摄像机标定. 自动化学报[J], 2002, 28(4): 488-496.
[45] 孟晓桥,胡占义. 摄像机自标定方法的研究与进展. 自动化学报[J],2003, 29(1): 110-124.
[46] 夏良正,李久贤. 数字图像处理. 东南大学出版社, 2005.
[47] 陆宗骐. C++ 图像处理编程. 清华大学出版社, 2005.
[48] 李晶皎,朱志良. 模式识别. 电子工业出版社, 2004.
[49] 朱虹. 数字图像处理. 科学出版社, 2005.
[50] Kenneth. R..Castleman . Digital image processing. 清华大学出版社, 2003.
[51] 谷口庆治. 数字图像处理. 科学出版社,2002.
[52] 董长虹等. Matlab 图像处理与应用. 国防工业出版社, 2004.
[53] 成礼智,王红霞等. 小波的理论与应用. 科学出版社, 2004.
[54] 黄爱民,安向京,骆力等. 数字图像处理与分析基础. 中国水利水电出版社, 2005.
[55] 张宏林,蔡锐等. VC++数字图像模式识别技术及工程实践. 人民邮电出版社, 2003.
[56] 龚建伟,熊光明. Visual C++ rbo C 串口通信编程实践. 电子工业出版社2004-10-1.
[57] 吴晓波. 应用多项式插值函数提高面阵 CCD 尺寸测量的分辨力. 《仪器仪表学报》,1996 / 17 / 02 P 154-158.
[58] 郭彦珍. 应用概率论提高 CCD 尺寸测量的分辨力. 《 仪器仪表学报》 , 1988 / 9 / 02 P 149-154.
[59] 邹仲力. 提高 CCD 尺寸测量分辨力的解调测量法[J]. 仪器仪表学报,1986,7(1):38~45.
[60] 韩军伟. 景物边缘提取技术的研究与应用. 西北工业大学硕士学位论文,2001 年 3 月:24～39.
[61] 张艳珍. 微机视觉相关理论及技术研究. 大连理工大学博士学位论文.2001 年 4 月:16～35.
[62] 王富治. 高精度视觉测量系统研究. 电子科技大学硕士学位论文. 2005.6.