Two-dimensional position and orientation tracking of micro-robot with a webcam

This paper proposes a 2D position and orientation tracking method of a micro-robot based on vision recognition. Position tracking is easy and simple, but the orientation of the micro-robot is hard to correctly recognize because the micro-robot is small and symmetrical. Since the orientation of the micro-robot is closely related to propulsion direction, it is necessary to recognize the orientation to accurately control the micro-robot. In this paper, the vision-based recognition result for orientation is compensated by expecting the direction of magnetic field from the Helmholtz coil currents.