Development and validation of YOLO models for water surface object detection in maritime surveillance

Effective water surface object detection is essential for maritime surveillance, enabling accurate identification of vessels, floating debris, and other potential hazards. This paper presents a detailed analysis of YOLO11, the latest advancement in the YOLO series of deep learning models, focusing exclusively on water surface object detection tasks. Incorporating advanced architecture, YOLO11 improves detection accuracy, robustness, and efficiency under variable lighting, weather, and sea states. Comprehensive evaluations using two maritime datasets show superior performance in detecting small, occluded, and dynamic objects compared to YOLOv8, YOLOv9 and YOLOv10, while maintaining real-time inference. The findings highlight YOLO11’s potential to enhance maritime surveillance, enabling safer and more efficient monitoring systems.