散射中心特征序贯匹配的SAR图像目标识别方法  

Object recognition in SAR imagery using sequential feature matching of scattering centers

作　　者：唐涛[1] 粟毅[1] 

机构地区：[1]国防科学技术大学电子科学与工程学院,湖南长沙410073

出　　处：《系统工程与电子技术》2012年第6期1131-1135,共5页Systems Engineering and Electronics

基　　金：国家自然科学基金(60972120)资助课题

摘　　要：在合成孔径雷达(synthetic aperture radar,SAR)图像目标识别领域,提取目标区分性强的识别特征一直是研究热点之一。设计了一种SAR图像目标识别的算法,利用目标散射中心点集特征和属性散射中心特征分量估计值序贯匹配识别车辆目标。实测数据的实验证实了算法在目标方位角变化情况下和俯仰角变化情况下,车辆目标的识别都有较好的表现。Extracting high discriminable features has been one of the focuses in synthetic aperture radar （SAR） automatic target recognition（ATR） research. An ATR method is proposed based on sequential feature matching of scattering centers. The feature is extracted from the point sets of scattering centers and is estimated by utilizing the attribute scattering center model. Then the sequential feature matching method is employed to determine the identity consistency of vehicle objects. Experiment results of real data with different azimuth angles and depression angles illustrate that the proposed method has a good performance.

关 键 词：合成孔径雷达 目标识别 特征匹配 属性散射中心 

分 类 号：TN957[电子电信—信号与信息处理]