基于机器视觉的高精度测量与装配系统设计
|
摘要
在对某些精密产品实现自动化生产过程中,存在难以对装配该产品所需的多种装配小零件进行高精度自动测量与装配的问题;针对该问题,搭建了基于机器视觉技术的自动化测量与装配系统;基于Halcon图像处理软件平台,对零件图像进行了中值滤波、图像增强等预处理;采用了Canny算法对零件求取像素精度的边缘,并运用椭圆曲线拟合法获取了亚像素精度边缘;建立了两种相机镜头畸变模型,采用径向排列约束(RAC)标定法与张正友标定法对相机进行了标定,并对标定精度进行了对比;实验结果表明:本系统的装配同轴度精度能达到0.05 mm,零件尺寸测量标准差低于3.8μm,满足工业需求,可以解决工业实际问题。
In realizing automation production process of certain precision products,there exist the problem of being hard to measure and assemble many kinds of small parts which are necessary for the product high-precision and autamatically.Aiming at the problem,the automatic measurement and assembly system is built based on machine vision technology.Based on Halcon,the image processing software platform,the system pretreats the parts image with median filtering,image enhancement and so on,and uses the Canny algorithm to calculate the pixel-precision edge,then uses elliptic curve fitting method to obtain the subpixel-precision edge.Two kinds of camera lens distortion model are established,using the method of the radial alignment constraint(RAO and the method of Zhang Zhengyou to calibrate cameras and compared the calibration accuracy.The experimental results show that the assembly alignment accuracy of system can reach 0.05 mm,and the standard deviation of measurement is less than 3.8 μm.It can meet the requirement of industry and solve practical problems.
In realizing automation production process of certain precision products,there exist the problem of being hard to measure and assemble many kinds of small parts which are necessary for the product high-precision and autamatically.Aiming at the problem,the automatic measurement and assembly system is built based on machine vision technology.Based on Halcon,the image processing software platform,the system pretreats the parts image with median filtering,image enhancement and so on,and uses the Canny algorithm to calculate the pixel-precision edge,then uses elliptic curve fitting method to obtain the subpixel-precision edge.Two kinds of camera lens distortion model are established,using the method of the radial alignment constraint(RAO and the method of Zhang Zhengyou to calibrate cameras and compared the calibration accuracy.The experimental results show that the assembly alignment accuracy of system can reach 0.05 mm,and the standard deviation of measurement is less than 3.8 μm.It can meet the requirement of industry and solve practical problems.
引文
[1]王冠英,胡心平,刁维龙.基于机器视觉的复杂形状模具尺寸测量[J].计算机测量与控制,2015,23(3):706-708.
[2]查英,刘铁根,杜东,等.基于机器视觉的零件自动装配系统[J].天津大学学报,2006,39(6):722-726.
[3]李钊宝,汪新新,贾茜媛.基于机器视觉的工件尺寸高精度测量方法[J].计算机测量与控制,2015,23(6):1866-1870.
[4]郭瑞,刘振国,曹云翔,等.基于视觉的装配机器人精确定位研究[J].制造业自动化,2014,36(5):154-156.
[5]Casten Steger,Markus Ulrich,Christian Wiedemann.Machine Vision Algorithms and Applications[M].lsted.,Yang Shaorong,Wu Dijing,Duan Deshan,translated.Beijing:Tsinghua University Press,2008:97-125.
[6]张丽群.基于相关法的多重成像亚像素定位算法的研究与实现[D].哈尔滨:哈尔滨工程大学,2005.
[7]解则晓,辛少辉,李绪勇,等.基于单目视觉的机器人标定方法[J].机械工程学报,2011,47(5):35-39.
[8]赵高长,武风波,周彬,等.基于DLT模型的摄像机标定简化方法[J].应用光学,2009,30(4):585-589.
[9]张娜.摄像机标定方法的研究[D].南京:南京理工大学,2014.
[10]Lanser S.Model-based localization based on monocular video images[D].Technical University of Munich,1997.
[11]Zhang Z Y.A flexible new technique for camera calibration[J].IEEE Transactions on Patten Analysis and Machine Intelligence,2000,22(11):1330-1334.
[2]查英,刘铁根,杜东,等.基于机器视觉的零件自动装配系统[J].天津大学学报,2006,39(6):722-726.
[3]李钊宝,汪新新,贾茜媛.基于机器视觉的工件尺寸高精度测量方法[J].计算机测量与控制,2015,23(6):1866-1870.
[4]郭瑞,刘振国,曹云翔,等.基于视觉的装配机器人精确定位研究[J].制造业自动化,2014,36(5):154-156.
[5]Casten Steger,Markus Ulrich,Christian Wiedemann.Machine Vision Algorithms and Applications[M].lsted.,Yang Shaorong,Wu Dijing,Duan Deshan,translated.Beijing:Tsinghua University Press,2008:97-125.
[6]张丽群.基于相关法的多重成像亚像素定位算法的研究与实现[D].哈尔滨:哈尔滨工程大学,2005.
[7]解则晓,辛少辉,李绪勇,等.基于单目视觉的机器人标定方法[J].机械工程学报,2011,47(5):35-39.
[8]赵高长,武风波,周彬,等.基于DLT模型的摄像机标定简化方法[J].应用光学,2009,30(4):585-589.
[9]张娜.摄像机标定方法的研究[D].南京:南京理工大学,2014.
[10]Lanser S.Model-based localization based on monocular video images[D].Technical University of Munich,1997.
[11]Zhang Z Y.A flexible new technique for camera calibration[J].IEEE Transactions on Patten Analysis and Machine Intelligence,2000,22(11):1330-1334.