This paper proposes a tracking controller for obstacle avoidance of an omni-directional mobile vehicle (OMV) using potential function method through a ceiling-mounted camera system. To do this task, the followings were done. At first, for image processing, a ceiling-mounted camera system was installed to get images. The goal point and obstacles were separated and recognized by a color recognition method. Second, potential function method was proposed to generate the shortest path to avoid obstacles from the start point to the goal point. The mobile platform is a mecanum wheel omni-directional vehicle. Third, the tracking controller is designed for the OMV to track a trajectory generated by a path planning algorithm based on the backstepping method using Lyapunov function. Finally, the simulation results were presented to show the effectiveness of the proposed trajectory planning algorithm and tracking controller for obstacle avoidance of the OMV.