Motion Path Control of Multi-DOF Handling Manipulator in Industrial Field Based on Vision and C-Space Obstacle Avoidance

Abstract

In order to realize the intelligentization of the multi-DOF handling manipulator in the industrial field, this paper presents a motion path control approach which combines the vision based identification and localization and C-space (Configuration space) based obstacle avoidance planning. The objects in the image captured by CCD camera can be extracted by gray processing, background subtraction, binarization, and edge detecting, the work pieces and the obstacles can be matched based on roundness and squareness, and then the coordinates of above objects can be obtained by parallax and coordinate conversion. In addition, both the source and destination coordinates in C-space can be obtained by mapping these configurations position and orientation into C-space, and the motion path can be automatically generated by variable step A* search in free area which is built based on solid manipulator model, then the handling work can be finished and obstacle avoidance can also be realized. The results of simulation and experiment demonstrate the scientificalness and effectiveness of this control approach.