基于三维机器视觉的工业机器人定位系统设计

Design of Industrial Robot Positioning System Based on 3D Machine Vision

为解决工业机器人在复杂自动化生产线上目标精确定位问题,文章搭建了由史陶比尔工业机器人、Kinect视觉传感器和上位机组成硬件平台,设计了基于三维视觉的工业机器人抓取定位系统。通过棋盘标定法对工业机器人抓取系统进行标定,采用深度边缘分割方法采集图像并提取工作面,根据累计概率Hough变换法提取工作面上不同形状目标的中心点,编写上位机处理程序实现图像分割、目标中心点定位及机器人运动控制等功能。实验结果表明该系统能够对规则形状目标实现定位功能,定位精度满足工业机器人的工作需求。

In order to solve the target precise positioning problem of industrial robots in complex automation production lines,a hardware platform based on Staubli industrial robot,Kinect vision sensor and the upper computer is set up and industrial robot grasping positioning system based on three-dimensional visual is designed.The system is calibrated by the checkerboard calibration method.Image is acquired and working plane is extracted by the depth edge segmentation method.The center points of different targets on the working plane are extracted by the cumulative probability Hough transform method.Image segmentation,target center location and robot motion control are realized by writing the upper computer processing program.The experimental results show that the system can realize the positioning function of rule shape grabbing targets and the positioning accuracy can meet the working requirements of industrial robots.