基于多分辨率能量特征的SAR目标检测  

Target Detection Method in SAR Imagery Based on Multiresolution Energy Features

作　　者：刘东平[1] 杨志国[1] 黄晓涛[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：《雷达科学与技术》2007年第6期451-454,共4页Radar Science and Technology

基　　金：国家自然科学基金(No.60402034)

摘　　要：针对UWB SAR中二面角目标和树干杂波在方向域和频率域的不同特性,特别是频率域相关特性的不同,给出了一种新的多分辨率划分方法,通过该方法对原始图像的处理得到多分辨率图像序列,并以该图像序列为对象提取能量特征,最终实现了目标检测。对实测数据的检测表明,该方法能有效地抑制杂波,提高目标的检测性能。Based on the different features of target and trunk clutter in azimuth and range in UWB SAR imagery, especially the different correlation characteristics in frequency domain, this paper presents a new muhiresolution image decomposition algorithm. Multiresolution image sequence can be obtained through transformation of raw images by use of this decomposition algorithm. And then energy features that finally realize target detection can be extracted through multiresolution image sequence. The experimental results indicate that the algorithm can suppress trunk clutter and improve the performance of target detection effectively.

关 键 词：超宽带合成孔径雷达 多分辨率 目标检测 能量特征 

分 类 号：TN957[电子电信—信号与信息处理] TN958[电子电信—信息与通信工程]