人工智能深度学习在海事监管系统的应用  

Application of Deep Learning Based on Artificial Intelligence in Maritime Supervision System

作　　者：张圣东 温启锐 Zhang Shengdong;Wen Qirui(Shenzhen Communication Center of Southern Navigation Service Center,Shenzhen,Guangdong 518000,China)

机构地区：[1]南海航海保障中心深圳通信中心,广东深圳518000

出　　处：《中国海事》2024年第5期15-17,共3页China Maritime Safety

摘　　要：水上事故和灾难不仅给人们的生命财产带来严重威胁,也对海洋环境造成了严重破坏。通过采用人工智能深度学习法,使用YOLOv8目标检测技术,结合摄像头开发了一套针对漂浮蚝排与其他海上相关要素的海事监管系统,最终得到模型的准确率为91.7%。同时讨论深度学习在海事监管系统的局限性,结果表明深度学习应用于海事监管系统切实有效。Maritime accidents and disasters not only pose a grave threat to people's lives and property but also cause severe harm to the marine environment.This paper introduces a maritime supervision system for floating oyster rafts and other searelated elements using the deep learning method of artificial intelligence(AI)and the YOLOv8 object detection technology with the camera.The accuracy of the final model is 91.7%.It also studies the limitations of deep learning when it is applied in the maritime administrative system.The results show that the application of deep learning in the maritime supervision system is an effective approach.

关 键 词：深度学习 海上危险物 智能识别 海事监管 YOLOv8 目标检测 

分 类 号：U698[交通运输工程—港口、海岸及近海工程]