Research on Fast Shaft Hole Assembly Technology Based on 2D Point Cloud Matching Shape Recognition Method by Industrial Robot

Abstract

Precise and rapid shaft hole assembly is important for improving equipment performance and assembly efficiency. As the variety of parts and shapes, the accuracy and efficiency of manual assembly are limited. In this paper, a shaft hole assembly method by using image recognition algorithm and robot is proposed. In the image recognition process, a shape recognition method based on 2D point cloud matching is presented, and combined with robot to realize grasping and assembly movement. In order to improve the recognition accuracy and reduce the real-time computational complexity, the method with binarization, Gauss filtering and shape recognition are provided to process the image obtained from camera. The experiment results show that the proposed image recognition could effectively identify part shape and complete the assembly process accurately, i.e., the method could effectively improve the efficiency and accuracy of shaft hole assembly.