System for Automatic Vessel Detection and Tracking in the Port of Split

This paper presents the development of an automatic vessel detection and tracking system in the Port of Split using computer vision and machine learning. The system enables the determination of key attributes such as vessel distance, size and speed. YOLOv8 is used for detection, an exponential function for distance approximation, and DeepSORT tracker for continuous tracking in video. The system has potential applications in maritime safety and violation detection