Infrared ship target detection method based on YOLOv5

Abstract

At present, the mainstream vessels automatic identification system mainly consists of shore based (base station) facilities and onboard equipment. These discriminate systems mainly rely on VHF and GPS, which cannot function properly under radio backdrop. This article recommend a vessels mark discriminate maneuver stem from infrared vivids. This maneuver uses the saturated emulate YOLOv5 algorithm. We used the recordset already established by Shandong University to establish the YOLOv5 blueprint. thread cultivate and testing, the divine accuracy on this recordset rapieceed 100%.