Detection and tracking for the awareness of surroundings of a ship based on deep learning

Abstract To prevent maritime accidents, it is crucial to be aware of the surrounding environment near ships. The images recorded by a camera mounted on ship could used for awareness other ships it. In this study, was performed using three procedures: detection, localization, and tracking. Initially, detection deep learning-based model, YOLO (You Only Look Once) v3, based image. A virtual image dataset constructed Unity overcome difficulty obtaining onboard with various sizes ships, improve performance. This followed localization procedure in which position horizon calculated orientation information from ship. Subsequently, detected spatial coordinate system information. Following this, position, course over ground, speed ground target were tracked time domain extended Kalman filter. learning model that determines heading proposed utilize abundant cameras, set initial value Finally, method validated an actual video captured installed encountering scenarios. tracking results compared automatic identification data obtained As result, entire procedures showed good performance, estimated surroundings ship, images, future.