Autonomous UAV Control for Maritime Applications using Deep Reinforcement Learning-based Image Optimisation

Yuanqing Yang; Emmanouil Spyrakos-Papastavridis; Mingfeng Wang; Yansha Deng

IROS 20250 citations

Autonomous UAV Control for Maritime Applications using Deep Reinforcement Learning-based Image Optimisation

Yuanqing Yang, Emmanouil Spyrakos-Papastavridis, Mingfeng Wang, Yansha Deng

Abstract

In this paper, we present an autonomous control system for Unmanned Aerial Vehicles (UAVs), specifically designed to inspect a detected suspicious vessel and capture information-rich images in a maritime environment. The maritime environment is ever-changing and uncertain, making it challenging to perform maritime monitoring tasks efficiently and reliably. The proposed UAV control system consists of multiple modules, including path planning, vessel searching, image processing, and image optimization. A novel image optimization approach utilizing deep reinforcement learning (DRL) is proposed to enhance the quality of the captured images by jointly controlling the movement of the UAV and camera orientation. The effectiveness and efficiency of the proposed system were validated and evaluated by searching the vessel and optimizing the captured images in the self-developed simulation environment in Gazebo.

BibTeX

@inproceedings{iros2025_autonomousuavcon,
  title = {Autonomous UAV Control for Maritime Applications using Deep Reinforcement Learning-based Image Optimisation},
  author = {Yuanqing Yang and Emmanouil Spyrakos-Papastavridis and Mingfeng Wang and Yansha Deng},
  booktitle = {IROS 2025},
  year = {2025}
}