Geometric adaptive dynamic visual servoing of a quadrotor UAV

Abstract

A geometric adaptive dynamic visual servoing controller is proposed to control a quadrotor UAV to the desired pose defined by a visual image of a planar target. Different from other existing works on visual servoing, this paper introduces the idea of geometric control into the adaptive dynamic visual servoing controller design for UAVs. Partial pose information is recovered as the feedback signal by using the homography between the current image and desired image. Geometric control, together with adaptive backstepping techniques, is employed to design the visual servoing controller with an online adaptation law for the unknown depth. Based on Lyapunov techniques, it is shown that the proposed adaptive controller guarantees asymptotic stability. Simulation results are provided to illustrate the performance of the proposed control scheme.