A Novel Hand-eye Calibration Method of a Robot-Laser Scanner System for Large-scale Metrology

Abstract

High-flexibility and high-precision are required in large-scale measurement for advanced manufacture. A robot-laser scanner system, which inherits the good flexibility of industrial robot and high accuracy of laser scanner, is a research focus recently. To obtain high measurement accuracy, hand-eye calibration is essential for this system. A novel two-step method, including primary calibration and optimization calibration, is proposed to calibrate the hand-eye of the measurement system. The primary calibration method is based on laser tracker and a well-designed tetrahedron, which can avoid the effect resulted from the parameter errors of industrial robot. Additionally, an optimizing calibration method by transforming the hand-eye calibration result into SE (3) is proposed to calculated the mean value of multiple primary calibration results, aiming to improve the calibration accuracy. Finally, a comparative experiment is conducted to verify the feasibility and accuracy of optimizing calibration method. It was found that comparing to the typical hand-eye calibration method, the optimizing method improves the mean repeat measurement accuracy by 23.4%, and the absolute measurement accuracy is also improved, with max/mean accuracy of 0.222 mm/0.178 mm.