基于散射机理分类与频谱相关性分析的极化SAR人造目标鉴别  

Man-made Target Discrimination Based on Scattering Mechanism Identification and Spectral Correlation Property Analysis of POLSAR Images

作　　者：徐牧[1] 王雪松[1] 肖顺平[1] 

机构地区：[1]国防科技大学电子科学与工程学院,长沙410073

出　　处：《信号处理》2009年第3期412-417,共6页Journal of Signal Processing

基　　金：新世纪优秀人才计划项目资助(NCET-04-0997);国家自然科学基金(60672033)

摘　　要：已有极化SAR人造目标鉴别方法通常仅利用较为简单的特征来区分目标与杂波,用于鉴别的信息量有限。针对这一不足,本文提出一种基于散射机理分类与频谱相关性分析的鉴别新方法。该方法充分利用人造目标与自然杂波在散射机理类型及频谱相关性等方面的差异进行目标鉴别,不同类型鉴别信息的综合利用使本文方法具有理想的鉴别效果。基于E-SAR全极化SAR实测数据验证了本文方法的有效性和优良性能。Traditional techniques of man-made target discrimination in polarimetric synthetic aperture radar （POLSAR） images usually make an unsatisfactory discrimination using very simple features and limited information. To overcome this shortcoming, a novel discrimination scheme based on scattering mechanism identification and spectral correlation property analysis is proposed to automatically discriminate man-made targets in diverse natural clutter environments of POLSAR images. Our novel method, which makes fully use of the difference between man-made target and natural clutter on scattering mechanism type and spectral correlation property, permits to obtain a clear and credible target discrimination effect. The validity of the novel method is verified by fully polarimetric SAR data acquired by DLR＇ s airborne E-SAR system.

关 键 词：人造目标鉴别 极化SAR 散射机理 频谱相关性 

分 类 号：TN958[电子电信—信号与信息处理]