3D Cameras and Algorithms for Multi-Angle Gripping and Control of Robotic Arm

Abstract

This research develops an automated multi-angle identification and gripping path planning method of a robotic arm. First, we integrate a 3D camera to obtain the image, position, and distance of the workpiece and then send the image to the remote host via a network connection to identify the workpiece and calculation path with a deep learning algorithm. Through the process, the best path and the angle of arm movement are found. The experimental results show that the system continuously reads real-time images from the 3D camera and performs the calculations to correct the moving path when the arm moves. The overall operation is very smooth, and the workpiece can be accurately clamped.