AUTONOMOUS TRACKING AND LANDING ON MOVING GROUND VEHICLE WITH MULTI-ROTOR UAV

In this manuscript, an autonomous system to track a moving ground vehicle and landing on it with multi-rotor UAV is proposed. Technical development of such system is discussed in detail. It includes sensors selection and integration, target detection algorithm and implementation, mathematical model of UAV and flight controller design. The system utilized near infrared camera which is able to detect the marker even at night or low illuminance, up to 18 Hz with processor on-board UAV. The complete system was first simulated in MATLAB, and then implemented to an actual UAV. The successful flight trials with a small size quad-rotor UAV landing autonomously on a moving small truck have shown that the design is valid and viable to actual applications. The proposed vision-laser target tracking performance has achieved a 99.2% success rate with a static marker, and a 94.4% success rate with a moving marker.