Autonomous Landing of a UAV on a Custom Ground Marker using Image-Based Visual Servoing

Landing is a crucial maneuver for a successful flight mission. This paper presents vision-based ground marker detection and autonomous landing of Unmanned Aerial Vehicles (UAVs). The methodology is based on a monocular vision system that detects the predefined ground marker and determines the altitude of the UAV solely from the vision sensor data. The same vision system is also used for positioning and estimating the orientation of the UAV with respect to the ground marker. A proportional controller is implemented for landing the UAV accurately. Image Based Visual Servoing (IBVS) control technique is implemented to reduce the computational cost of the system. The algorithm accounts for abnormal situations such as loss of ground marker in the image frame or a partially visible ground marker. The performance of the landing system is verified through both simulations and outdoor experimentation. The results show the accuracy and robustness of the proposed autonomous landing strategy.