SAR satellites, unaffected by illumination and weather conditions, enable round-the-clock and all-weather earth observation, offering numerous distinct advantages in monitoring ships and their tracks on the sea surface. Therefore, this paper selects images from three C-band radar satellites, namely GF-3, GF-3B, and GF-3C, to create a remote sensing dataset for ship wake recognition. The YOLOv8-OBB model is trained and utilized for the identification of ship wakes. Experimental results demonstrate that, compared to the YOLOv8 model, YOLOv8-OBB exhibits superior recognition performance, making it more suitable for obtaining motion information such as ship positions and headings. The recognition accuracy reaches 86%. This method can be applied to tasks such as maritime ship monitoring, holding significant importance for assessing maritime economic and trade activities, ensuring energy transportation, and safeguarding sea surface security.
Ship wake target detection in C-SAR image based on YOLOv8-OBB
International Conference on Remote Sensing Technology and Image Processing (RSTIP 2024) ; 2024 ; Dali, China
Proc. SPIE ; 13640
2025-06-03
Conference paper
Electronic Resource
English
Ship target detection based on CBAM-YOLOv8
SPIE | 2024
|Ship Detection Based on Improved YOLOv8 Algorithm
IEEE | 2024
|ITD-YOLOv8: An Infrared Target Detection Model Based on YOLOv8 for Unmanned Aerial Vehicles
DOAJ | 2024
|Vehicle Target Detection on Campus Roads Based on YOLOv8
IEEE | 2024
|