Visual servo control of unmanned aerial vehicles: An object tracking-based approach

This paper investigates the object following control problem for a low-cost quadrotor unmanned aerial vehicle (UAV). An object tracking-based approach is proposed. In this approach, the techniques of object tracking and visual servo control are efficiently combined to achieve the control objectives. To eliminate the impact of tracking target rotation during tracking model, an online learning method is proposed by learning the object size over different view at the desired distance. On the basis of this, an image-based visual servo control law is then designed to ensure that the quadrotor can track the moving target in real-time and keep the desired distance away from the object. The applicability of the proposed scheme is verified by flight experiments. It is demonstrated that the proposed algorithm is robust to rotary movement.