摘要
如何在复杂环境中准确、快速地实现各类船舶目标检测是一项重要研究课题,也是保障船舶航行安全、保护领土安全、加强水域资源监管的基础。红外成像、合成孔径雷达(Synthetic Aperture Radar,SAR)以及卫星遥感等成像技术的发展为船舶检测提供了丰富的图像数据,已取得许多研究成果。介绍了船舶目标检测的过程,从基于传统图像处理技术的船舶目标检测和基于深度学习的船舶目标检测两方面总结并分析了现有船舶目标检测方法,讨论了相关关键技术,最后指出了未来的研究方向。
How to detect all kinds of ship targets accurately and quickly in complex environment is an important research and ship detection technology is also the foundation of ensuring the navigation safety of ships,protecting territorial security and strengthening the supervision of water resources.The development of remote sensing technologies such as infrared imaging,synthetic aperture radar(SAR)and satellite remote sensing provides abundant image data for ship detection and many research findings based on remote sensing images have been achived.This paper introduces the ship detection process,summarizes and analyzes the existing ship detection methods based on traditional image processing technology and deep learning,discusses related key technology,and finally points out the future research direction.
作者
王伟
WANG Wei(Southwest China Institute of Electronic Technology,Chengdu 610036,China)
出处
《电讯技术》
北大核心
2020年第9期1126-1132,共7页
Telecommunication Engineering
基金
国家重点研发计划项目(2017YFC1404900)。
关键词
船舶检测
遥感图像
图像处理
深度学习
ship detection
remote sensing image
image processing
deep learning
