Intelligent Operation of Robot for Handling and Assembly Based on Vision Guidance

Abstract

In spite of the wide use of industrial robot for handling and assembly nowadays, the assembly of heavy workpieces is still a challenge due to the weight of the object and requirements of high reliability and precision. For this problem, a multi-camera guidance system based on monocular vision is designed, which uses the complementary information from different cameras. The system architecture of combining “eye-in-hand” and “eye-to-hand” is adopted. A solution to recognition and positioning of the workpiece based on machine vision and improved Hough Transform is provided. A scheme of automatic rectification and searching for the best position and orientation between the workpiece and track beam is proposed. Finally, based on the KUKA KR16 industrial robot, the experiment of handling and assembly is designed and accomplished to verify the effectiveness of the system.