A full 6 degree-of-freedom visual estimation with laser assisted

Abstract

Efficient and accurate measurement in unstructured environment is a challenging issue and also research interest in the industrial field. The accuracy of vision is very high in the close-range scene and the stability of laser measurement is better, with a large achievable range. A 6-DOF pose measurement method based on vision and assisted by laser is proposed in this paper. By combining their respective advantages, the method can be immune to ambient light. The optimization objective is established by using the projection coordinates of laser spots in different coordinate systems and solved by Newton–Raphson iteration. The initial value is estimated by Kronecker product. The Monte Carlo method is further used to simulate. The causes of measurement errors and the influence factors affecting the accuracy are analyzed. On this basis, an experimental prototype is constructed. The accuracy and repeatability accuracy experiments are conducted to verify the correctness of the proposed method.