Overhead tracking and control of a mobile robot

An issue with bio-inspired robotics is knowledge transfer from biology to robots. Our paper provides a tracking and robot controller framework to allow a robot to copy the movement behaviours of a fish. This method receives fish coordinates from an image and translates these as way-points to the robot, by utilising an over-head depth camera. Sensor-motor data and front-facing camera images for the robot will be provided to a Learning by Demonstration (LbD) model to understand the movement and decision making of the fish. As of this paper, initial data collection has been completed and a robot tracking algorithm has been developed.