Landing on a moving target using image-based visual servo control

Abstract

This paper addresses the problem of landing a Vertical Take-Off and Landing (VTOL) quadrotor using image-based visual servo control. Observable features on a flat target plane are exploited to derive a suitable control law. The kinematics of a VTOL quadrotor system are expressed in terms of an unnormalized spherical centroid of features and optical flow is used as velocity measurement for control. The dynamics with respect to the force input are dependent on the distance between the camera and the target plane, which is not measurable by a simple vision system. The proposed control law guarantees convergence without estimating any parameter related to the unknown distance, which is also guaranteed to remain strictly positive. Simulation and experimental results are provided to illustrate and assess the performance of the proposed controller.