A visual tracking system for UAV landing on ships
Bruno Damas, Nuno Pessanha Santos, Matilde Correia Vieira · 2024
Unmanned Aerial Vehicles (UAVs) have significantly widened the range of possible applications due to their unmanned capabilities. UAVs are commonly utilized in Intelligence, Surveillance, and Reconnaissance (ISR) operations, enabling them to gather important real-time data. To ensure operational reliability, performing the landing operation autonomously is crucial. However, performing an autonomous landing on a ship is challenging, as it is intended to be executed in an outdoor environment, over a moving platform, where good coordination and precise position estimation of the UAV position relative to the ship is crucial. We present a ship-based binocular vision system with temporal filtering based on an Extended Kalman Filter (EKF) to track the position of a UAV during landing. The developed system can detect noise and failures in one or both cameras, enabling accurate detection and guidance in these scenarios. The preliminary results show that this simple setup can perform tracking with low errors, demonstrating its potential for real-world applications.