开放式组焊数控系统的研制及焊缝识别的数字图像处理研究
|
摘要
我国汽车工业正面临着巨大的挑战，汽车大型覆盖件模具技术是车身制造技术的重要组成部分，也是形成汽车自主开发能力的一个关键环节。研制智能组焊机床则是是模具自动化生产的重要一环。
本文阐述了该装置-智能组焊机床的研制，详细介绍了其开放式结构的控制软件系统，说明了NC嵌入PC型数控系统的开发方式及重点、难点，如运动控制卡DDA工作原理、多线程技术、实时控制技术、数据库技术。在本文中，着重介绍了控制系统中焊缝识别的步骤和关键算法，涉及数字图像处理技术中的多个方面，包括畸变校正、图像拼接、边缘检测、直线提取等。本文实现并详细介绍了采用两次双线性插值法进行图像畸变校正的算法，对多种边缘检测算法的特点和适用范围进行了分析和比较，从中选取适合本项目的算法，提出并实现了基于特征像素点线段的匹配算法、金字塔分层处理的接触匹配算法，用于快速地进行图像的拼接；基于金字塔分层原理的接触匹配算法具有算法速度快，抗干扰能力强，无需知道图片的大概重合位置等特点。对直线提取算法Hough变换所存在的问题如线的连通性、断点的检测、计算量大等问题提出了相应的改进算法。采用以上算法，本文依据拍摄的模具图像对焊缝进行识别，并计算其坐标，根据实践检验，这是一种可行的方法。
在理论分析和研究的基础上，使用Visual C++ 6.0设计并实现了一个开放式的智能焊接数控系统，能在人工辅助操作下进行半自动化的焊缝示教，并可通过图像处理进行焊缝识别，自动批量地进行焊接，大大提高了生产率，成功应用于生产实践。
Chinese automotive industry is facing great challenges nowadays, the technology of automobile good-size moulds manufactures is an important part of automobile body manufacture technology, it also is a key of the ability of independent development in automotive production .To development a intelligent automatic welding equipment is a important part of moulds automatic manufactures.
This paper explains the research on welding equipment, introduces Open Architecture Numerical Control software System to you , analyses the method of development. This paper also discusses the difficulties and the emphases in development of the PC embedded by NC, such as the DDA operating theorem of motion control card, multi-threaded technology and realization of real-time and database technology. It lies a emphasis on the steps and some key algorithms concerned with recognition of welding Line, such as Geometric Correction , Image Mosaic , Straight Line Extraction and Edge Detection. This paper introduces the Geometric Correction algorithm , which uses bilinear interpolation twice . Have analyzed their characters and applicatilities, we select the fittest one to the project among the several kinds of methods in Edge Detection. Two methods, the one is a new matching algorithm based on feature pixel-lined segment, the other is contact-match algorithm based on hierarchal pyramid are presented, you can utilize
them in the Image Mosaic. Algorithm based on pyramid has a coarse-to-fine process. Fast speed , better anti-jamming and avoid knowing approximate matching area are the main good features of the algorithm. In the aspect of detection of the straight line, this thesis introduces Hough transform, which has been advanced in some of following aspect compared with it's general algorithm: the judgment of the connectivity of line , the detection of broken point, the complexity of algorithm. According above algorithm, we extract the welding lines from image and calculate their coordinates, it was proved a feasible way.
On the base of theory analysis and research, I implemented a opened architecture numerical control welding system using Visual C++, which can detect welding lines assisted by person or extract welding lines from image. We can implement task of weld automatically and batchly,which lead to promote the productivity effectively, so it was put to use in industry.
本文阐述了该装置-智能组焊机床的研制，详细介绍了其开放式结构的控制软件系统，说明了NC嵌入PC型数控系统的开发方式及重点、难点，如运动控制卡DDA工作原理、多线程技术、实时控制技术、数据库技术。在本文中，着重介绍了控制系统中焊缝识别的步骤和关键算法，涉及数字图像处理技术中的多个方面，包括畸变校正、图像拼接、边缘检测、直线提取等。本文实现并详细介绍了采用两次双线性插值法进行图像畸变校正的算法，对多种边缘检测算法的特点和适用范围进行了分析和比较，从中选取适合本项目的算法，提出并实现了基于特征像素点线段的匹配算法、金字塔分层处理的接触匹配算法，用于快速地进行图像的拼接；基于金字塔分层原理的接触匹配算法具有算法速度快，抗干扰能力强，无需知道图片的大概重合位置等特点。对直线提取算法Hough变换所存在的问题如线的连通性、断点的检测、计算量大等问题提出了相应的改进算法。采用以上算法，本文依据拍摄的模具图像对焊缝进行识别，并计算其坐标，根据实践检验，这是一种可行的方法。
在理论分析和研究的基础上，使用Visual C++ 6.0设计并实现了一个开放式的智能焊接数控系统，能在人工辅助操作下进行半自动化的焊缝示教，并可通过图像处理进行焊缝识别，自动批量地进行焊接，大大提高了生产率，成功应用于生产实践。
Chinese automotive industry is facing great challenges nowadays, the technology of automobile good-size moulds manufactures is an important part of automobile body manufacture technology, it also is a key of the ability of independent development in automotive production .To development a intelligent automatic welding equipment is a important part of moulds automatic manufactures.
This paper explains the research on welding equipment, introduces Open Architecture Numerical Control software System to you , analyses the method of development. This paper also discusses the difficulties and the emphases in development of the PC embedded by NC, such as the DDA operating theorem of motion control card, multi-threaded technology and realization of real-time and database technology. It lies a emphasis on the steps and some key algorithms concerned with recognition of welding Line, such as Geometric Correction , Image Mosaic , Straight Line Extraction and Edge Detection. This paper introduces the Geometric Correction algorithm , which uses bilinear interpolation twice . Have analyzed their characters and applicatilities, we select the fittest one to the project among the several kinds of methods in Edge Detection. Two methods, the one is a new matching algorithm based on feature pixel-lined segment, the other is contact-match algorithm based on hierarchal pyramid are presented, you can utilize
them in the Image Mosaic. Algorithm based on pyramid has a coarse-to-fine process. Fast speed , better anti-jamming and avoid knowing approximate matching area are the main good features of the algorithm. In the aspect of detection of the straight line, this thesis introduces Hough transform, which has been advanced in some of following aspect compared with it's general algorithm: the judgment of the connectivity of line , the detection of broken point, the complexity of algorithm. According above algorithm, we extract the welding lines from image and calculate their coordinates, it was proved a feasible way.
On the base of theory analysis and research, I implemented a opened architecture numerical control welding system using Visual C++, which can detect welding lines assisted by person or extract welding lines from image. We can implement task of weld automatically and batchly,which lead to promote the productivity effectively, so it was put to use in industry.
引文
[1]左飞雁，张桂香．基于Windows的开放式数控系统的实现．计算技术与自动化，2003，(9)：85-87
[2]丁仕燕，韩江．开放式数控系统软件平台的研究．机械与电子，2002(1)153-155
[3]彭礼孝．虚拟设备驱动程序开发起步与进阶．北京：人民邮电出版社，2000：30-45
[4]Chris Cant．Writing Windows WDM Device Drivers．北京：机械工业出版社，2000．1：117-119
[5]PCI-8136M 6-Axis Motion Controller Card User's Guide. ADLINK Technology Inc. Manual Rev. 1.00: Oct 20, 2000
[6]王润孝．机床数控原理与系统．西安：西北工业大学出版社，1993：133-135
[7]潘爱民．Visual C++技术内幕 (第四版)．北京：清华大学出版社，1999：45～56．245～266
[8]Jim Beveridge & Robert Wiener著．侯捷译．Win32多线程程序设计．武汉：华中科技大学出版社，2002．1：76-80
[9]李现勇．Visual C++串口通信技术与工程实践．人民邮电出版社，2002．5：223～236
[10]刘刀桂，孟繁晶编著．Visual C++实践与提高—数据库篇．北京：中国铁道出版社，2001．3：163-170
[11]Kenneth R．Castlema著．朱志刚等译．数字图像处理(新版)．北京：电子工业出版社，2002．2：412～422
[12]阮秋镝．数字图像处理学．北京：电子工业出版社，2001．1：273～279
[13]姜大志．计算机视觉中三维重构的研究与应用．博士论文，南京航空航天大学，2002：32～35
[14]马颂德，张正友．计算机视觉——计算理论与算法基础．北京：科学出版社，1998：55～59，88～92
[15]Z.Jericevic, D.M.Benson, J. Bryan, and L.C.Smith. Geometric Correction of Digital Images Using Orthonormal Decomposition. Journal of Microscopy, 1987,149(3):233-245
[16]J. R. Bergen, P. Anandan, K. J. Hanna, and R. Hingorani. Hierarchical model-based motion estimation. In Second European Conference on Computer Vision(ECCV'92),
Santa Margherita Liguere, Italy, Springer-Vedag, 1992:237～252
[17]C.-H. Lee and L.-H. Chen. A Fast Motion Estimation Algorithm Based on the Block Sum Pyramid. IEEE Trans. Image Processing, 1997, 6:1587～1591
[18]游素亚，杨静．图像边缘检测技术的发展与现状．新技术，1999，5(4)：7～75
[19]韦春桃，程晓宇．LOG算子进行边缘检测的研究．桂林工学院学报，2001，20(3)：83～85
[20]R. Nevatia, K. Babu. Linear Feature Extraction and Description. Computer Graphics Image Processing, 1980,13(3):257-269
[21]J.B. Bums, A.R. Hanson and EM. Riseman. Extractigng Straight Lines. JEEE Transactions on Pattern Analysis and Machine Intelligence, 1986,4:425-455
[22]M.Boldt, R. Weiss and E.Riseman. Taken-based extraction of straight lines. IEEE Trans. Syst.Man Cybem, 1989,19:1581-1594
[23]P.V.C. Hough. Method and Means for Recognizing Complex Patterns. U.S. Patent 3,069,654.Dec. 18,1962
[24]何斌，马天予，王运坚，朱红莲．Visual C++数字图像处理(第二版)．北京：人民邮电出版社，2002：235-255
[25]R. Deriche. Using Canny's Criteria to Derive a Recursively Implemented Optimal Edge Detector. International Journal of Computer Vision, 1987,1(2): 10-17
[26]J.C.Bezdek, R.Chandrasekhar, Y.Attikiouzel. A Geometric Approach to Edge Detection. IEEE Transactions on Fuzzy System, 1994,4(1):52-75
[27]H.Li, M.A.Lavin, and R. J.Le Master. Fast Hough Transform: A Hierarchical Approach,CVGIOP, 1986,36:139～161
[28]V.J.Milenkovic. Mltiple Resolution Search Techniques for the Hough Transform in High Dimension Parameter Spaces. Techniqes for 3D Machine Perception, North-Holland, 1986:231-254
[29]J.Illingworth and J.Kittlr. The Adaptive Hough Transform. IEEE Trans. Pattern Anal.Machine Intell, 1987,9(5):690-698
[30]毓晋．图像工程．北京：清华大学出版社，1999：182-183
[31]张杏．低信噪比图像中多运动目标的实时识别研究与应用．硕士研究生论文，湖南大学，2003：31～34
[32]周心明，兰赛，徐燕．图像处理中几种边缘检测算法的比较．现代电力，2000．8：56～59
[33]周道炳，朱卫纲．几种边缘检测算子的评估．指挥技术学院学报，2001，8(3)：87～93
[34]徐士良．常用算法程序集．北京：清华大学出版社，1996．11：23～26
[35]Kenneth R．Castleman著．朱志刚，石定机等译．数字图像处理(新版)．北京：电子工业出版社，2002．2：214～220
[36]张宏林编著．Visual C++数字图像模式识别技术及工程实践．北京：人民邮电出版社，2003．1：319～325
[37]杨枝灵，王开等编著．Visual C++数字图像获取、处理及实践应用．北京：人民邮电出版社，2003．1：393～399
[38]文贡坚，王润生．从航空遥感图像中自动提取主要道路．软件学报，2001，11(7)：957～964
[39]J.M.Jolion, A.Rosenfeld. An O(log n)Pyramid Hough ransform. Pattern Recognition Letters, 1989,(9):343-349
[40]J.Illingworth, J.Kittler. A Survey of the Hough Transform. Computer Vision. Graphics and Image Processing, 1988, 44(1):87-116
[41]L.Carrioli, M.Diani and L.Lombardi. Hierarchical Hough:Advantages and Automatic Construction of Models. Intern.Conf.On Pattern Recognition, 1988:791-794
[42]文贡坚．航空图像中道路和建筑物的自动提取技术．博士学位论文．国防科学技术大学，2000．4：24～25
[43]钟力，胡晓峰．重叠图像拼接算法．中国图像图形学报，1998，3(5)：32～36
[44]李志刚，纪玉波等．边界重叠图像的一种快速拼接算法．计算机工程，2000，5(26)：77～81
[45]廖士中，高培焕，苏艺，王大鹏．一种光学镜头摄像机图像几何畸变的修正方法．中国图像图形学报，2000，5：97～101
[2]丁仕燕，韩江．开放式数控系统软件平台的研究．机械与电子，2002(1)153-155
[3]彭礼孝．虚拟设备驱动程序开发起步与进阶．北京：人民邮电出版社，2000：30-45
[4]Chris Cant．Writing Windows WDM Device Drivers．北京：机械工业出版社，2000．1：117-119
[5]PCI-8136M 6-Axis Motion Controller Card User's Guide. ADLINK Technology Inc. Manual Rev. 1.00: Oct 20, 2000
[6]王润孝．机床数控原理与系统．西安：西北工业大学出版社，1993：133-135
[7]潘爱民．Visual C++技术内幕 (第四版)．北京：清华大学出版社，1999：45～56．245～266
[8]Jim Beveridge & Robert Wiener著．侯捷译．Win32多线程程序设计．武汉：华中科技大学出版社，2002．1：76-80
[9]李现勇．Visual C++串口通信技术与工程实践．人民邮电出版社，2002．5：223～236
[10]刘刀桂，孟繁晶编著．Visual C++实践与提高—数据库篇．北京：中国铁道出版社，2001．3：163-170
[11]Kenneth R．Castlema著．朱志刚等译．数字图像处理(新版)．北京：电子工业出版社，2002．2：412～422
[12]阮秋镝．数字图像处理学．北京：电子工业出版社，2001．1：273～279
[13]姜大志．计算机视觉中三维重构的研究与应用．博士论文，南京航空航天大学，2002：32～35
[14]马颂德，张正友．计算机视觉——计算理论与算法基础．北京：科学出版社，1998：55～59，88～92
[15]Z.Jericevic, D.M.Benson, J. Bryan, and L.C.Smith. Geometric Correction of Digital Images Using Orthonormal Decomposition. Journal of Microscopy, 1987,149(3):233-245
[16]J. R. Bergen, P. Anandan, K. J. Hanna, and R. Hingorani. Hierarchical model-based motion estimation. In Second European Conference on Computer Vision(ECCV'92),
Santa Margherita Liguere, Italy, Springer-Vedag, 1992:237～252
[17]C.-H. Lee and L.-H. Chen. A Fast Motion Estimation Algorithm Based on the Block Sum Pyramid. IEEE Trans. Image Processing, 1997, 6:1587～1591
[18]游素亚，杨静．图像边缘检测技术的发展与现状．新技术，1999，5(4)：7～75
[19]韦春桃，程晓宇．LOG算子进行边缘检测的研究．桂林工学院学报，2001，20(3)：83～85
[20]R. Nevatia, K. Babu. Linear Feature Extraction and Description. Computer Graphics Image Processing, 1980,13(3):257-269
[21]J.B. Bums, A.R. Hanson and EM. Riseman. Extractigng Straight Lines. JEEE Transactions on Pattern Analysis and Machine Intelligence, 1986,4:425-455
[22]M.Boldt, R. Weiss and E.Riseman. Taken-based extraction of straight lines. IEEE Trans. Syst.Man Cybem, 1989,19:1581-1594
[23]P.V.C. Hough. Method and Means for Recognizing Complex Patterns. U.S. Patent 3,069,654.Dec. 18,1962
[24]何斌，马天予，王运坚，朱红莲．Visual C++数字图像处理(第二版)．北京：人民邮电出版社，2002：235-255
[25]R. Deriche. Using Canny's Criteria to Derive a Recursively Implemented Optimal Edge Detector. International Journal of Computer Vision, 1987,1(2): 10-17
[26]J.C.Bezdek, R.Chandrasekhar, Y.Attikiouzel. A Geometric Approach to Edge Detection. IEEE Transactions on Fuzzy System, 1994,4(1):52-75
[27]H.Li, M.A.Lavin, and R. J.Le Master. Fast Hough Transform: A Hierarchical Approach,CVGIOP, 1986,36:139～161
[28]V.J.Milenkovic. Mltiple Resolution Search Techniques for the Hough Transform in High Dimension Parameter Spaces. Techniqes for 3D Machine Perception, North-Holland, 1986:231-254
[29]J.Illingworth and J.Kittlr. The Adaptive Hough Transform. IEEE Trans. Pattern Anal.Machine Intell, 1987,9(5):690-698
[30]毓晋．图像工程．北京：清华大学出版社，1999：182-183
[31]张杏．低信噪比图像中多运动目标的实时识别研究与应用．硕士研究生论文，湖南大学，2003：31～34
[32]周心明，兰赛，徐燕．图像处理中几种边缘检测算法的比较．现代电力，2000．8：56～59
[33]周道炳，朱卫纲．几种边缘检测算子的评估．指挥技术学院学报，2001，8(3)：87～93
[34]徐士良．常用算法程序集．北京：清华大学出版社，1996．11：23～26
[35]Kenneth R．Castleman著．朱志刚，石定机等译．数字图像处理(新版)．北京：电子工业出版社，2002．2：214～220
[36]张宏林编著．Visual C++数字图像模式识别技术及工程实践．北京：人民邮电出版社，2003．1：319～325
[37]杨枝灵，王开等编著．Visual C++数字图像获取、处理及实践应用．北京：人民邮电出版社，2003．1：393～399
[38]文贡坚，王润生．从航空遥感图像中自动提取主要道路．软件学报，2001，11(7)：957～964
[39]J.M.Jolion, A.Rosenfeld. An O(log n)Pyramid Hough ransform. Pattern Recognition Letters, 1989,(9):343-349
[40]J.Illingworth, J.Kittler. A Survey of the Hough Transform. Computer Vision. Graphics and Image Processing, 1988, 44(1):87-116
[41]L.Carrioli, M.Diani and L.Lombardi. Hierarchical Hough:Advantages and Automatic Construction of Models. Intern.Conf.On Pattern Recognition, 1988:791-794
[42]文贡坚．航空图像中道路和建筑物的自动提取技术．博士学位论文．国防科学技术大学，2000．4：24～25
[43]钟力，胡晓峰．重叠图像拼接算法．中国图像图形学报，1998，3(5)：32～36
[44]李志刚，纪玉波等．边界重叠图像的一种快速拼接算法．计算机工程，2000，5(26)：77～81
[45]廖士中，高培焕，苏艺，王大鹏．一种光学镜头摄像机图像几何畸变的修正方法．中国图像图形学报，2000，5：97～101
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |