Simultaneous calibration of hand-eye relationship, robot-world relationship and robot geometric parameters with stereo vision

Abstract

Robot-world calibration and hand-eye calibration are fundamental steps for robots equipped with cameras. This problem is described as solving “AX = YB” equation. However, this problem usually interacts with robot calibration, because robot geometric parameters are not very precise. Therefore, we propose a novel calibration method which can calibrate hand-eye relationship, robot-world relationship and robot geometric parameters simultaneously. This method considers hand-eye relationship and robot-world relationship as another two axes of robot arm. Modified D-H convention is used to describe robot position and orientation and stereo cameras are used to measure positioning errors. In order to improve calibration accuracy, least square method is introduced to solve calibration equations. Simulation and implement prove that this overall calibration method is feasible.