计算机工程

一种基于MLS的SAR图像点特征提取方法

展开
  • 1.沈阳航空航天大学 计算机学院,沈阳110136;
    2.辽宁大学 信息学院,沈阳110036
石祥滨(1963-),男,辽宁沈阳人,教授,主要研究方向:图像处理,虚拟现实,E-mail:sxb@sau.edu.cn。

收稿日期: 2013-03-16

基金资助

国家自然科学基金(项目编号:61170185);辽宁省科技攻关计划项目(项目编号:2011217002)

A method of SAR image point feature extraction based on MLS

Expand
  • 1.School of Computer Science,Shenyang Aerospace University,Shenyang 110136;
    2.College of Information,Liaoning University,Shenyang 110036

Received date: 2013-03-16

摘要

针对合成孔径雷达图像点特征提取问题,提出了一种基于MLS的SAR图像点特征提取方法。首先利用提出的基于超像素关联分析的SAR图像分割算法得到二值分割图像,再把该二值图像和SAR图像进行点乘运算,获取到含有强度信息的目标区域,然后采用移动最小二乘法(MLS)对目标区域进行曲面拟合,根据设定的点特征判决规则,最后提取出SAR图像的峰值、脊和谷等多种点特征。基于MSTAR数据的实验结果表明了该方法的有效性和准确性。

本文引用格式

石祥滨,刘进立,张劲松,陈润锋 . 一种基于MLS的SAR图像点特征提取方法[J]. 沈阳航空航天大学学报, 2013 , 30(3) : 38 -43 . DOI: 10.3969/j.issn.2095-1248.2013.03.009

Abstract

In order to solve the problem of the SAR image point feature extraction,this article proposes a method of SAR image point feature extraction based on MLS.Firstly,the article proposes a SAR image segmentation algorithm based on super-pixel correlation analysis.The algorithm is used to obtain segmentation result.Point multiplication is executed between the binary image and SAR image to generate the target area containing the intensity information.Then,the moving least square method is used to fit the surface based on discrete pixels in the target area.Finally,according to point feature decision rule,the method can extract multiple features of SAR image,such as its peak value,ridges,valleys,etc.Experimental results based on the data of MSTAR show the validity and accuracy of the proposed method.

参考文献

[1]计科峰,匡纲要,粟毅,等.SAR图像目标峰值特征提取与方位角估计方法研究[J].宇航学报,2004,25(1):102-108.
[2]高贵,计科峰,匡纲要,等.高分辨率SAR图像目标峰值提取及峰值稳定性分析[J].电子与信息学报,2005,27(4):561-565.
[3]Szeliski R,Zabih R,Scharstein D,et al.A comparative study of energy minimization methods for markov random fields with smoothness-based priors[J].IEEE Traction on Pattern Analysis and Machine Intelligence,2008,30(6):385-394.
[4]Bae M H,Pan R,Wu T,et al.Automated segmentation of mouse brain images using extended MRF[J].NeuroImage,2009,46(3):717-725.
[5]刘向华,周荫清,孙慕涵.基于快速退火MRF的改进SAR图像分割方法[J].北京航空航天大学学报,2010,36(6):719-722.
[6]Achanta R,Shaji A,Smith K,et al.Slic superpixels[R].EPFL Technical Report no.149300,2010.
[7]马苗,田红鹏,张艳宁.灰色理论在图像工程中的应用研究[J].中国图象图形学报,2007,12(11):1943-1951.
[8]程玉民.移动最小二乘法研究进展与述评[J].计算机辅助工程,2009,18(2):5-11.
[9]Manson J,Schaefer S.Moving least squares coordinates[J].Computer Graphics Forum,2010,29(5):1517-1524.
[10]张翠,郦苏丹,王正志.基于MRF场的SAR图像分割方法[J].遥感技术与应用.2001,16(1):66-68.
[11]Robert A,Weisenseel W,Clem K,et al.Markov random field segmentation methods for SAR target chips[J].SPIE,1999,3721:102-108.
文章导航

/