舰船目标识别技术研究进展  

Advances in ship target recognition technology

作　　者：马啸 邵利民[1] 金鑫[1] 徐冠雷[1] MA Xiao;SHAO Limin;JIN Xin;XU Guanlei(Department of Navigation,Dalian Naval Academy,Dalian 116018,China)

机构地区：[1]海军大连舰艇学院航海系

出　　处：《科技导报》2019年第24期65-78,共14页Science & Technology Review

基　　金：国家自然科学基金项目(61471412,61771020)

摘　　要：舰船目标的有效识别和监控对维护海洋权益、保障海上航行安全至关重要。根据舰船目标信息的获取形式,从辐射噪声信号、雷达回波信号、卫星遥感图像、合成孔径雷达图像、红外图像、可见光图像几个舰船目标的主要信息获取来源出发,阐述了舰船目标识别技术的研究进展,总结分析了目前基于不同信号源的舰船目标识别方法普遍存在的具有高度任务相关性、计算成本高与运行时间长等问题。结合深度学习技术在语音识别、图像识别等领域的发展,建议将基于深度学习技术的典型目标识别方法Faster R-CNN及YOLO引入舰船目标识别领域,以研究鲁棒性更好、准确率更高、实时性更强的舰船目标识别方法。The effective identification and monitoring of ship targets is essential for safeguarding the maritime rights and ensuring the navigation safety.In line with the acquisition form of the ship target information,this paper reviews the ship target recognition technology based on several main information acquisition sources of the ship targets,including the radiated noise signal,the radar echo signal,the satellite remote sensing image,the synthetic aperture radar image,the infrared image and the visible image.The current research difficulties in the ship target recognition methods based on different signal sources are analyzed,involving the high mission correlation,the high calculation cost and the long running time.Combined with the development of the deep learning technology in the speech recognition,the image recognition and other fields,the typical target recognition methods based on the deep learning technology,the Faster R-CNN and the YOLO,are applied in the ship target recognition.It is proposed that the introduction of the deep learning technology into the ship target recognition field indicates a new direction for the research of the ship target recognition methods with better robustness,higher accuracy and better real-time performance.

关 键 词：舰船 目标识别 深度学习 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] U675.79[自动化与计算机技术—计算机科学与技术]