AN ADAPTIVE SCHEME FOR IMAGE-BASED VISUAL SERVOING OF AN UNDERACTUATED UAV

Abstract

An image-based visual servoing (IBVS) method is proposed for controlling the 3D translational motion and the yaw rotation of a quadrotor. The dynamic model of this unmanned aerial vehicle (UAV) is considered at the design stage to account for its underactuation. In contrast with previous IBVS methods for underactuated UAVs, which used spherical image moments as visual features, the proposed controller makes use of appropriately defined perspective moments. As a consequence, we gain a clear improvement in performance, as satisfactory trajectories are obtained in both image and Cartesian space. In addition, an adaptation mechanism is included in the controller to achieve robust performance in spite of uncertainties related to the depth of the image features and to the dynamics of the robot. Simulation results in both nominal and perturbed conditions are presented to validate the proposed method.