Simultaneous calibration of hand-eye and kinematics for industrial robot using line-structured light sensor

Abstract

For visually guided industrial robots, it is necessary to calibrate the robot's hand-eye relationship to accurately guide the robot's operations with the measurement data of the visual coordinate system. In most cases, the calibration process of hand-eye error does not consider the impact of kinematic errors, which accumulates the error in the calibration process and reduces the calibration accuracy. To enhance the calibration accuracy of industrial robots, a simultaneous calibration method of robot hand-eye relationship and kinematics using a line-structured light sensor is proposed. A simultaneous calibration model is established by using the differential kinematics principle. Afterward, the line-structured light sensor and two standard balls with known center distances are used in the calibration and accuracy verification experiments. After calibration and compensation with the proposed method, the average distance error of the robot declined from 3.967 mm to 1.001 mm. The proposed simultaneous calibration method for hand-eye relationship and kinematics dramatically improves the robot's positioning accuracy.