Improvement on vision guidance in AUV docking

Autonomous Underwater Vehicle docking technology is now extensively used in the field of underwater. Vision guidance, as one of its core components, shows a great development prospect. In this paper, a brief description of the vision guidance models and the previous algorithm were described. Both the advantages and disadvantages of that algorithm would be analyzed. By using grey-scale feature analysis, edge detection and morphology methods, we attempted to improve the algorithm of calculating the centers of target lights, getting both robustness and accuracy of vision guidance system improved, which would make further mechanical control easier.