图像视觉特征的机械臂末端位姿监测方法研究

Research on the Pose Monitoring Method of Robotic Arm End Based on Image Visual Features

为解决当前监测方法位姿误差和高度误差偏大等问题,提出图像视觉特征的机械臂末端位姿监测方法。先通过TSAI手眼标定法标定机械臂手眼系统,然后采用粗定位提取机械臂轮廓,利用最小二乘法对边缘点拟合,获取视觉特征,最后通过机械臂视觉特征获取抓取位置,对抓取位姿和关节角进行映射得到机械臂作业轨迹,实现机械臂末端位姿监测。仿真实验结果表明,所提方法的位姿误差仅为0.06mm,高度误差仅为0.05mm,运行时间也较短,明显优于对比方法,不仅可有效降低位姿误差和高度误差,还能有效提升监测效率。

In order to solve the problems of large pose error and height error of the current monitoring method,a method for moni-toring the pose of the robot arm end based on image visual features is proposed.First by TSAI hand-eye calibration mechanical arm hand-eye calibration system,and then USES the coarse location to extract the contour,mechanical arm using the least squares fitting to the edge point,visual characteristics,at last,through visual characteristics for fetching position,mechanical arm to grab posture and joint Angle of the mapping get mechanical arm operation trajectory,realize the mechanical arm end posi-tion monitoring.The simulation results show that the pose error of the proposed method is only 0.06mm,the height error is only 0.05mm,and the running time is also shorter,which is obviously better than the comparison method.It can not only effectively re-duce the pose error and height error,but also effectively Improve monitoring efficiency.