摘要
针对当前光学成像卫星的目标识别基本依靠人工判别的现状,结合中低分辨率遥感图像的目标特征不明显、舰船种类多、识别困难的特点,论文根据国内卫星资源情况与研究现状,分析并比较了光学成像卫星不同传感器提取的目标特征,选择可见光传感器进行目标识别方法研究,对适用于中低分辨率遥感图像的目标识别方法进行比较,并选择出适用于当前海军海上侦察预警的方法。该方法能够极大缩减卫星图像数据处理时间,改善舰艇目标识别由人工判别的现状,增加战时卫星数据的及时性,为海军侦察预警提供技术支撑。
According to the current status of ship target recognition based on optical imaging satellite that mainly rely- on artifi- cial, combined with the characteristic of medium-low resolution remote sensing image as unobvious characteristics of target, vari- ous types of ships, and difficulty- to recognize, this paper analyzes the domestic satellite resources and research quo, compares and chooses the characteristic extraction method between different sensors which applies for navy maritime reconnaissance. This method can reduce the processing time of image data, change the status quo of artificial interpretation, increase the timeliness of wartime satellite data, and provide technical support for navy reconnaissance and scout.
作者
刘星璇
饶世钧
洪俊
LIU Xingxuan;RAO Shijun;HONG Jun(Department of Postgraduate Management,Dalian Navy-Academy,Dalian 116018;Department of Information System,Dalian Navy-Academy,Dalian 116018)
出处
《舰船电子工程》
2018年第6期21-25,共5页
Ship Electronic Engineering
基金
军委科技委前沿科技创新项目(编号:17-H863-05-ZT-001-003-02)资助
关键词
中低分辨率
光学成像卫星
遥感图像
舰船目标识别
medium-low resolution
optical imaging satellite
remote sensing image
ship target recognition
