Planar marker-based localization of a robotic fish in a public aquarium

A robotic fish localization system in a public aquarium is presented in this paper. The system is built on the marker-based camera pose estimation method. A webcam, an ARM processor, and multiple planar markers are employed for real-time camera pose estimation. The camera is attached at the bottom of the head of the robotic fish. Since the camera and the fish body are connected rigidly, the pose estimation method gives the absolute 6 degree of freedom (DOF) location of the robotic fish. The experiments above ground and underwater validate the proposed localization system. Through the tests in a public aquarium with the integrated robotic fish system, the autonomous swimming was successfully demonstrated.