To allow incorporation of autonomous Unmanned Aerial Vehicles (UAV's/drones) into maritime military operations, it is critical to be able to accurately localize the UAV with respect to the moving maritime vessel during the take-off and landing phases. This work addresses the study and implementation of a visual detection, tracking and threedimensional positioning method for a specific drone from a moving maritime vessel. The YOLOv5 detector and the OceanPlus tracker have been trained on a custom dataset with good performance in accuracy and processing time. The drone's position with respect to the vessel is estimated by applying stereo triangulation to the centres of the bounding boxes returned by the object detectors and trackers. The performance of the proposed positioning method was evaluated in a realistic simulated environment in the Unreal Game Engine. The proposed method allows detection, tracking, and positioning of a target drone at ranges exceeding 100m while achieving positioning errors below 10cm during landing phases.

Ship deck landing of Unmanned Aerial Vehicles (UAVs/drones) in different kinds of environmental conditions remains a bottleneck for the widespread deployment of UAVs for maritime operations. For safe operation, the relative motion between the UAV and the pitching and rolling deck of a moving ship must be estimated accurately and in real-time. This paper presents a visual Simultaneous Localization and Mapping (SLAM) method for real-time motion estimation of the UAV with respect to its confined landing area on a maritime platform during landing phase. The visual SLAM algorithm ORB-SLAM3 [1] was selected after benchmarking with multiple state-of-the-art visual SLAM and Visual Odometry (VO) algorithms with the EuRoC dataset [2]. It was evaluated for a simulated landing scenario of a UAV at 16m height with a downward camera in multiple configurations with sufficient results in both speed and accuracy for the landing task.