首页 | 本学科首页   官方微博 | 高级检索  
     

基于稀疏先验的SAR图像目标方位角稳健估计方法
引用本文:郭炜炜,杜小勇,胡卫东,郁文贤. 基于稀疏先验的SAR图像目标方位角稳健估计方法[J]. 信号处理, 2008, 24(6)
作者姓名:郭炜炜  杜小勇  胡卫东  郁文贤
作者单位:国防科技大学ATR重点实验室,湖南,长沙,410073
基金项目:系国家重点实验室基金资助项目;项目
摘    要:
稳健的高精度目标方位角估计能有效提高SAR ATR的计算效率和识别性能.SAR图像中目标的近雷达主导边界包含较为精确的目标方位角信息,可用于目标方位角估计.由于目标电磁散射特性以及SAR图像斑点噪声的影响,提取的目标近雷达主导边界很不规则,存在"野值"点.本文根据"野值"点稀疏分布的特性,利用最大后验原理提出了一种稳健的方位角估计方法.该方法能够有效检测和剔除主导边界中的"野值",从而提高目标方位角估计的精度和稳健性.针对仅利用距离主导边界估计带来的目标垂直与水平方位的模糊性,基于分割图像中目标区域长宽比特征提出了一种解模糊的新方法.MSTAR实测数据的实验结果表明提出的算法具有较高的精度与稳健性.

关 键 词:方位角估计  稀疏分布  主导边界  长宽比

A Robust Target Aspect Estimation Method from SAR Images Based on Sparse Prior
GUO Wei-wei,DU Xiao-yong,HU Wei-dong,YU Wen-xian. A Robust Target Aspect Estimation Method from SAR Images Based on Sparse Prior[J]. Signal Processing(China), 2008, 24(6)
Authors:GUO Wei-wei  DU Xiao-yong  HU Wei-dong  YU Wen-xian
Abstract:
In SAR ATR systems,the computation efficiency and recognition performance will be greatly improved if the target as- pect is estimated accurately and robustly in advance. The leading edge pixels closest to the radar have more aspect information,but it is commonly irregular due to the target scattering characteristic and speckle noise. According to the sparse distribution characteristic of the outlier pixels in the leading edge,a robust target aspect estimation algorithm based on MAP principle is presented to detect and eliminate the outlier pixels and estimate the target aspect subsequently. Moreover,a new method is introduced to eliminate the ambiguity of hori- zontal and vertical aspect estimation based on the length-width ratio of the target region. Experimental results with the MASTAR data show the validity and robustness of the algorithm.
Keywords:SAR ATR
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号