The Grasping Point For Planar Workpiece Based On Fuzzy Connectedness Prior Knowledge

This paper describes the problem that a manipulator grasps the planar workpiece with a negative pressure suction cup. The grasping point must be determined to meet the command of planar grasping criterion to lay the foundation of the workpiece placement. The first contribution of this paper is the segmentation of the workpiece under industry condition with illumination and occlusion in the background. The method of iterative graph cuts and relative fuzzy connectedness are combined to segment the workpiece. The second contribution is the computation of the optimal grasping point using point cloud of the workpiece. The point cloud coming from the feature points, which are extracted from the structured light stripes according to the range of the image segmentation and transformed to the coordinates of the base, is used to segment the largest plane of the workpiece which is feasible for grasping. The normal vector of the planar workpiece are also determined. The experiment results show that the algorithm is effective and feasible in grasping the planar workpiece.