Development of vision-based autonomous robotic fish and its application in water-polo-attacking task

This paper presents the design and locomotion control of a vision-based autonomous robotic fish, and its application in water-polo-attacking task. Most of previous work on task strategies of autonomous robots is focused on the terrestrial robots and seldom deals with underwater applications. In fact, the tasks in underwater environment are more challenging than those in ground circumstances due to the uncertainties and complexity in hydro-environment. In this paper, a water-polo-attacking task is designed and implemented based on artificial potential field. The experimental results are presented to verify the effectiveness of the proposed control algorithms.