A Novel Rotational Laser Vision Sensor for Industrial Robot Calibration

Industrial robots require high-precision calibration to perform tasks with accuracy and reliability in real-world environments. This study presents a novel rotational laser vision sensor designed to enhance the calibration accuracy of industrial robots. The key contribution of this paper is the high-precision modeling and calibration method for the proposed sensor, which effectively improves accuracy and repeatability. The sensor combines a rotating laser scanning mechanism with a monocular vision system, offering a compact design, high-efficiency area scanning, and laser resistance to interference. The sensor’s performance was evaluated through a series of experiments using a four-sphere calibration object for robot calibration. The Levenberg-Marquardt optimization algorithm was applied to identify the robot’s Denavit-Hartenberg (DH) parameters and external transformation matrices. The results show a substantial reduction in positioning errors, with the maximum error decreasing from 2.7468 mm to 0.1345 mm and the standard deviation from 0.5629 mm to 0.0296 mm after calibration.