基于机器视觉圆环零件尺寸提取方法研究  

Research on the dimensional extraction method of ring parts based on machine vision

作　　者：刘润浩 兰孝文[1] 王少锋[2] 李若飞 LIU Runhao;LAN Xiaowen;WANG Shaofeng;LI Ruofei(Information Engineering School,Inner Mongolia University of Science and Technology,Baotou 014010,China;Mechanical En-gineering School,Inner Mongolia University of Science and Technology,Baotou 014010,China;Changsha Shaoguang Semiconduc-tor Co.,Ltd,Shenyang Branch,Shenyang 110034,China)

机构地区：[1]内蒙古科技大学信息工程学院,内蒙古包头014010 [2]内蒙古科技大学机械工程学院,内蒙古包头014010 [3]长沙韶光半导体有限公司沈阳分公司,辽宁沈阳110034

出　　处：《内蒙古科技大学学报》2023年第3期271-276,共6页Journal of Inner Mongolia University of Science and Technology

基　　金：内蒙古自治区科技计划资助项目(2020GG0160).

摘　　要：针对流水线生产作业中圆环零件尺寸难以高精度快速提取的问题,提出了一种基于机器视觉的非接触式测量方法.通过机器人末端夹持线激光扫描仪扫描零件并将扫描数据转换到基座标下得到原始点云图像,对点云图像进行去噪处理和分离内、外圆特征点,提出了改进的随机采样一致性算法,分别拟合圆环内、外圆并计算圆环同心度.结果表明:该方法提取的圆环尺寸与真值比较误差在0.1 mm以内,满足自动化高精度测量需求.A non⁃contact measurement method based on machine vision was adopted to solve the problem of that the dimensions of ring parts in assembly line production operations is difficult to extract accurately and quickly.The point cloud image was obtained by the ro⁃bot's end clamped line laser scanner and the scanning data was converted to the base mark to obtain the original point cloud image.The point cloud image was denoised and the inner and outer circles were separated.An improved random sampling consistency algorithm was proposed to fit the inner circle and outer circle respectively and calculate the concentricity of the ring.The results show that the er⁃ror between the ring size extracted by this method and the true value is less than 0.1 mm,which meets the requirement of automatic high precision measurement.

关 键 词：激光扫描 机器视觉 随机采样一致性 圆环零件尺寸 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]