An Intelligent Grasping Platform Based on Point Clouds

Zijie Zhang ； Jing Zeng

《World Scientific Research Journal》 9卷4期 (2023/04) Pp. 177-184

https://doi.org/10.6911/WSRJ.202304_9(4).0025

摘要

An intelligent grasping platform based on point clouds is designed to execute effective manipulator grasping for clustered objects. For the software, ROS(Robot Operating System) is used as the programming framework to achieve the trajectory planning and control, and PCL(Point Cloud Library) is used to process point cloud data. For the hardware, X‐Arm6 6‐DOF manipulator and DAHUAN AG‐95 gripper are combined to construct a robot arm, and Intel RealSense D435i is fixed at the end of the arm to collect point cloud data in real time. Several algorithms are verified on the proposed platform. Experiment results show that all the selected methods can be well applied to the designed platform, and the algorithm combining improved GPD and point cloud registration trick performs best.

關鍵字

Grasping ； Point clouds ； ROS ； Clustered objects