多波段SAR图像隐蔽和暴露目标标注  

Hidden and Exposed Targets Labeling in Multi-Band SAR Images

作　　者：刘向君[1] 饶伟[1] 杨小小[1] 

机构地区：[1]海军潜艇学院,山东青岛266071

出　　处：《雷达科学与技术》2008年第4期278-281,共4页Radar Science and Technology

摘　　要：为了实现对场景的全面把握,为目标精确打击提供依据,文中提出并研究了多波段SAR图像中隐蔽和暴露目标的标注方法,该方法综合利用了基于决策级融合的多波段SAR图像目标检测和多波段SAR图像分类等方法。将该方法应用于P、L、X和Ku四个波段SAR图像,实现了对同一场景中的7辆卡车状态的标注。实验结果表明,该方法不但可以将场景中7辆卡车的隐蔽和暴露状态准确标注出来,而且能够准确显示目标原方位和姿态以及场景的细节特征,对目标攻击系统具有一定的辅助作用。To provide data for precision attacking system and observe the war scene more clearly, this paper proposes and studies a labeling method of hidden and exposed targets in multi-band SAR images. This method integrates multi-band SAR target detection based on fusion and multi-band SAR image classification, etc. In the paper, the method is applied to P, L, X and Ku-band SAR images and labels the conditions of seven trucks in the same scene. Experiment result proves that this method can not only label the two conditions（hidden and exposed） of the targets exactly, but also display the locations and stances of the targets and the details of the scene, which is useful to target attacking system.

关 键 词：多波段SAR 目标标注 目标检测 决策级融合 图像分类 

分 类 号：TN958[电子电信—信号与信息处理]