ABSTRACT
Ship allision is a crucial topic for offshore wind farms (OWFs) due to catastrophic consequences. To safeguard the OWF facilities, operators adopt CCTV to detect and monitor passing ships in the waters near OWFs manually. This paper proposed an automatic ship detection system for offshore wind farms on the basis of machine vision technologies. In the system, a YOLOv3 algorithm is used to track and identify ships. A stereo vision algorithm is applied to locate ship positions. The system is tested in a real scenario and validated by comparing the visual location results and automatic identification system data. Extensive experiments demonstrate that the detection system of ships can automatically and accurately detect and locate ships. The potential uses of the system include ship invasion and allision avoidance.
