摘要
为了解决工业机器人对带有复杂纹理的平面型工件的识别与定位问题,将机器视觉技术与工业机器人相结合。利用改进的FREAK算法对目标工件进行快速识别,并根据工件的形状和单应矩阵计算出工件的形心坐标和旋转角度,再结合双目视觉系统,获取机器人抓取点的三维位姿信息,引导SCARA机器人抓取目标工件。实验结果表明:在复杂的工业环境下,机器人抓取系统能够准确识别出目标工件,具有良好的鲁棒性,同时也满足工业生产的实时性要求。
In order to solve the problem of recognition and localization of planar workpieces with complex texture by industrial robots,machine vision technology was combined with industrial robots. An improved FREAK algorithm was used to quickly identify the target workpiece. And according to the shape of the workpiece and the homography matrix,the centroid coordinates and rotation angle of the workpiece were calculated. Then obtained the three-dimensional position information of the grasping point of the robot with the binocular vision system,and the SCARA robot was guided to grasp the target workpiece. The experimental results show that in complex industrial environment,the robot grasping system can accurately identify the target workpiece,has good robustness,and also meets the real-time requirements of industrial production.
作者
党宏社
吴晨刚
孙俊龙
DANG Hong-she;WU Chen-gang;SUN Jun-long(School of Electrical and Information Engineering,Shaanxi University of Science and Technology,Xi'an 710021,China)
出处
《科学技术与工程》
北大核心
2019年第28期184-188,共5页
Science Technology and Engineering
基金
陕西省社会发展科技攻关项目(2015SF275)资助
作者简介
第一作者:党宏社(1962-),男,汉族,陕西武功人,博士,教授。E-mail:danghs@sust.edu.cn。
