Vision-Based Dynamic Positioning System Design and Experiment for an Underwater Vehicle

Abstract

In this paper, a vision-based control system is developed for a remotely operated underwater vehicles. We utilized four underwater lights as the visual markers to control the depth, horizontal, position and yaw of the underwater vehicle. The system design and control algorithms are described in details. This proposed visual control system is tested by pool experiments, in which the underwater vehicle succeeds in keeping position, depth and yaw using the PID control algorithms.