Vision-based Autonomous Landing Control of a Multi-rotor Aerial Vehicle on a Moving Platform with Experimental Validations

Abstract

This paper studies the problem of landing a multi-rotor unmanned aerial vehicle (UAV) on a moving platform (i.e., surface craft, car) based on visual servo. An efficient method is proposed to get the landmark position, in which the camera can follow the moving landmark to supply the relative position information between UAV and the platform. To avoid the divergence of the collected image caused by the swinging camera, an inertial measurement unit (IMU) is used to get the attitude of camera, which will provide the rectification information for the collected image. By using the collected image and the compensation information, we could obtain the proper image and then the relative position can be expressed through the principle of monocular ranging. To demonstrate the effectiveness of the proposed landing method, we present and analyze the experimental results, showing that the landing procedure is steady with a high accuracy.