摘要
针对液压千斤顶装配中人工上料方式存在的生产效率低下问题,设计了一套基于双目线激光视觉引导的上料系统。在分析缸筒和活塞杆的机械结构后,设计了机器人专用夹爪。在视觉识别过程中,利用SAD(Sum of Absolute Differences)算法实现工件的相似性度量,以经典的ICP(Iterative Closest Point)算法实现点云配准,从而获得模板点云与场景点云之间的旋转平移矩阵。根据预先示教的抓取点位姿坐标,结合相机标定结果,获得目标工件抓取点的实际位姿坐标,从而引导机器人实现缸筒和活塞杆的自动上料。
In view of the low production efficiency of manual feeding in the assembly process of hydraulic jack,a feeding system based on binocular laser vision guidance is designed.By analyzing the mechanical structure of cylinder and piston rod,the special clamping claw for robot is designed.In the process of visual recognition,the classical SAD(Sum of Absolute Differences)algorithm is used to measure the similarity of workpieces,and the ICP(Iterative Close Point)algorithm is used to register the point cloud and obtain the rotation translation matrix.Finally,according to the position and posture coordinates of the grasping point taught in advance,and combined with the camera calibration results,the actual position and posture coordinates of the grasping point of the target workpiece are obtained,so as to guide the robot to realize the automatic feeding of the cylinder barrel and piston rod.
作者
陈远方
魏新生
李保强
CHEN Yuanfang;WEI Xinsheng;LI Baoqiang(Industrial and Intelligent Center,SIPPR Engineering Group Co.,Ltd.,Zhengzhou 450007,China)
出处
《成组技术与生产现代化》
2022年第1期28-32,62,共6页
Group Technology & Production Modernization
关键词
液压千斤顶
双目线激光视觉
模板匹配
自动上料
hydraulic jack
binocular laser vision
template matching
automatic feeding
