一种基于线扫描相机的手眼标定技术  

A Hand-eye Calibration Technique Based on Line-scan Camera

作　　者：兰浩[1] 张曦[1] 尚继辉 Lan Hao;Zhang Xi;Shang Jihui

机构地区：[1]上海大学机电工程与自动化学院,上海200072 [2]航天制造科技有限责任公司,上海200233

出　　处：《计量与测试技术》2019年第5期7-10,15,共5页Metrology & Measurement Technique

基　　金：国家自然科学基金项目(51205243);国家重点研发计划课题(2016YFC0302402)

摘　　要：提出一种以固定点变位姿的方法标定线激光扫描相机与机器人的手眼关系,该方法分为两步标定出手眼矩阵,采用直径已知的标准球作为固定点靶标,首先控制机器人以平移的方式运动至少3次,然后机器人任意位姿运动至少4次;每次运动后用机器人末端执行器上安装的线激光扫描器扫描靶球,获取靶球表面轮廓上一列点云,对点云进行空间圆拟合,进而求出靶球的球心空间坐标;最后根据约束关系可列出方程,用最小二乘法或者奇异值分解求解手眼关系。实验结果表明:该方法快捷有效计算量较小,标定精度在0.2mm左右,可以用于线激光扫描相机与机器人现场手眼标定。This paper presents a method to calibrate the hand-eye relationship between a line laser scanning camera and a robot,the method is divided into two steps.Standard balls with knowndiameters are used as fixed targets.Firstly,the robot is controlled to move at least 3 times in the way of translation,and then moves at least 4 times in any pose of the robot.After each movement,a line laser scanner mounted on the end-effector of the robot was used to scan the target ball and obtain a row of point clouds on the surface contour of the target ball.The spatial coordinates of the center of the target ball were obtained by fitting the point cloud.Finally,the equation can be listed according to the constraint relation,and the hand-eye relation can be solved by least square method or singular value decomposition.Experimental results show that this method is fast and effective,less computation and higher precision,and can be used for line laser scanning camera and robot field hand-eye calibration.

关 键 词：手眼标定 线结构光 机器人 最小二乘法 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]