A novel cable-grasping planner for manipulator based on the operation surface. (February 2022)

• Record Type: Journal Article
• Authors: Wu, Wei; Zhu, Yiman; Zheng, Xianjie; Guo, Yu
• Is Part Of: Robotics and computer-integrated manufacturing. Volume 73(2022)
• Publication Date: 2022-02
• Subjects: Cable pose measurement -- Symmetric edge feature -- Operation surface -- Two-phase path planner -- Motion directional manipulability; Robots, Industrial -- Periodicals; Computer integrated manufacturing systems -- Periodicals; Robotics -- Periodicals; Robots industriels -- Périodiques; Productique -- Périodiques; Robotique -- Périodiques; 670.285
• DOI: 10.1016/j.rcim.2021.102252
• Languages: English
• ISSNs: 0736-5845

Abstract

Abstract: The manipulation of deformable linear objects is a hot spot and difficult problem in robotics research. Cables, as one kind of deformable linear objects, are characterized by elongation and uncertain shape. Although previous research has paid attention to deformable linear object's manipulation, it still remains a challenge to detect and grasp cables in unstructured environments. In this paper, a new cable-grasping method framework by analysis of cable characteristics is proposed. The framework consists of cable pose measurement and cable-grasping path planning. In the cable pose measurement mission, we design a symmetric edge feature according to the cable image projection principle. Then, a semantic segmentation based edge detection technique and a center line fitting method based on sorting by principal axis are applied to measure the cable pose accurately. In the cable-grasping path planning mission, the grasping pose is determined considering the force directional manipulability. The cable operation surface is constructed after the characteristic analysis of the cable, on which basis, a two-phase path planner is proposed, with the second phase planner BG-RRT guided by motion directional manipulability to further optimize the planning path. The performance of the approach is verified by experiments in both the simulation environment and the actual environment. Highlights: Proposing a symmetric edge feature based on the cable image projection principle. Designing an edge detection method based on semantic segmentation as well as a center line fitting method based on sorting by principal axis. Constructing the cable operation surface according to the characteristics of cable. A two-phase path planner based on operation surface for reducing collisions. A new BG-RRT algorithm guided by motion directional manipulability for further improving the efficiency of cable grasping.