Vision-Guided Tracking and Emergency Landing for UAVs on Moving Targets

Abstract: This paper presents a vision-based adaptive tracking and landing method for multirotor Unmanned Aerial Vehicles (UAVs) to ensure safe recovery during propulsion system failures that reduce the UAV’s maneuverability and responsiveness. Addressing challenges such as external disturbances and agile target movements, the research employs a modified adaptive pure pursuit guidance technique with an additional adaptation parameter to account for reduced maneuverability. Furthermore, an adaptive landing strategy is introduced that minimizes off-target landings caused by lateral tracking errors through lateral offset-dependent vertical velocity control. The system utilizes vision-based tag detection to determine the relative position of an Unmanned Ground Vehicle (UGV). Extensive simulations and testing demonstrate the effectiveness of the system in mid-mission emergency landing scenarios, enhancing the safety and reliability of UAV operations in dynamic environments.

Download paper here