A bearing-only control law for stable docking of unicycles

This paper proposes a new control method for stabilising control design for docking unicycle-like vehicles based on bearing-only information. An omni-directional panoramic camera is used to detect visual targets around the docking station and provides bearing (or heading) data for each observed landmark. The convergence of the controlled system is fully analysed and simulations are provided to demonstrate the ideal behaviour of the system. A robust and computationally cheap blob detection algorithm is proposed and results are provided to demonstrate its performance in extracting targets from cluttered scenes. Experimental results are presented demonstrating the performance of the algorithm on the ANU Nomadic Technologies Nomad XR4000 robot.