An algorithm of hand-eye calibration for arc welding robot

Abstract

Vision-based seam tracking sensor is now widely used in arc welding robot due to its non-contact, high- precision, real-time characteristics. It is able to detect the weld seam position, and transform the position from camera coordinates to robot base coordinates. To realize such transformation, the hand-eye matrix which is the transformation relation between the camera and end-effector (EE) of robot is needed. In this paper, the mathematical model of the hand-eye problem is given, and a hand-eye calibration algorithm is proposed. The simulation experiments are given to analyze the calibration errors caused by the measuring noise. The experiment results show that the proposed algorithm meets the demands of accuracy for the arc welding robot.