LAUSR

Abstract Three-dimensional(3-D) shape measurement for large-scale objects based on line structured light combined with industrial robot has been increasingly important because of its attributes of high-flexibility and high-robustness. To acquire the shape of large-scale objects, the line structured light obtained 3-D lines are stitched together by using robot hand-eye calibration results. However, in the line structured light technique, the traditional light plane calibration is complex and the result 3-D measurement accuracy is relatively low. Besides, because the light plane equation of the line structured light is arbitrary, the traditional standard ball hand-eye calibration method cannot be directly used. In this paper, a light plane calibration method is proposed by using binocular cameras, which simplifies the calibration process and improves the 3-D measurement accuracy. Furthermore, a modified space circle fitting method is proposed for the standard ball hand-eye calibration, which can be flexibly combined with the line structured light for arbitrary light plane. Based on the above improvement of the light plane calibration and the robot hand-eye calibration, the proposed measurement technique can accurately measure 3-D shape for large-scale objects. Theoretical analysis and experiments are provided to evaluate its performance.