Vessel Target Detection in Marine Environment Based on YOLOv5
HAN Qindong
SHI Fan
JIA Chen
ZHAO Meng
Abstract：Automation and intelligence are the trend in the marine sector,and object detection algorithms are central to achieving intelligence.Among the current object detection models,YOLOv5 is widely used in various fields due to its small size,fast running speed and easy deployment.However,YOLOv5 is not implemented for the marine environment and cannot achieve good performance when dealing with data in the marine domain.To make YOLOv5 can be better applied in marine environments,this paper analyses the data characteristics of marine environments and proposes three data pre-processing algorithms,such as dark-ening enhancement,fog enhancement and rain enhancement for these characteristics,so that the trained model can better cope with the complex environment and weather conditions in marine environments.After using those algorithms,the mean average precision(mAP)of the YOLOv5 model for predicting ship targets in the marine environment improves from 61.1%to 67.6%.
Keywords：object detectioncomputer visiondeep learning
Publication Date：2025-05-20
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：6( 1273-1278 )