3D Eye-to-Hand Coordination for Uninstructed Robot Grasp Planning

2022 12th International Conference on CYBER Technology in Automation, Control, and Intelligent Systems (CYBER)(2022)

引用 0|浏览1
暂无评分
摘要
Robot grasping is widely explored in factories, where most objects are instructed with rigid grasp planning. However, for disordered and cluttered objects, it is rarely applied due to difficult object recognition and robot trajectory constraints, which may bring low-cost performance. Normally, the shape and pose of an object, which can be represented as point cloud are preliminary feedback for robot grasping strategy. Therefore, the point cloud segmentation of disordered and clustered objects is critical for grasping manipulation tasks with hand eye coordination. This paper presents an algorithm for locating and dividing different parts by point cloud image in chaotic environment, to determine the reasonable grasping position of parts and the trajectory planning of manipulator. With the experiment on a eye-to-hand robot platform, the results show that the proposed algorithm is robust to different parts stacked and has a relatively better recognition performance.
更多
查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络