The Research of Container Grasping System Based on Edge Detection and Delaunay Triangulation Doubling

Abstract

Addressing the issues of low picking efficiency and the difficulty in improving the success rates of intelligent cargo box robots, an intelligent logistics cargo box detection and control system based on edge detection and Delaunay triangulation is proposed. Edge detection technology is integrated with a grasping mechanism to enable the boundaries of cargo boxes to be identified from images, followed by precise shape reconstruction and spatial analysis using Delaunay triangulation. The system is capable of extracting the edges of cargo boxes in various positions and orientations, thereby providing reliable real-time detection. Experimental results demonstrate that the average error between the cargo box tilt angle calculated by the algorithm and the actual angle is approximately 3°, thereby meeting the grasping requirements of the robot. Additionally, the grasp success rate was improved by 8.84%, and grasp time was reduced by 1.9 seconds compared to conventional grasping algorithms.