全部 标题 作者
关键词 摘要

OALib Journal期刊
ISSN: 2333-9721
费用:99美元

查看量下载量

相关文章

更多...

基于马尔可夫随机场的SAR图象目标分割

DOI: 10.11834/jig.200208255

Keywords: 图象目标分割,SAR,马尔可夫随机场,合成孔径雷达,目标识别,目标获取

Full-Text   Cite this paper   Add to My Lib

Abstract:

运动、静止目标获取与识别(MSTAR)计划表明,将合成孔径雷达(SAP)图象分割成目标、阴影和背景杂波区域对于从开放环境中进行目标识别是一种有效的手段。但是由于SAP图象所固有的斑点噪声的影响,传统的分割方法很难获得准确的分割。为此提出了一种基于MRF(Markovrandomfield)模型的SAP图象分割算法。用MRF模型描述待分割图象的先验知识,利用最大似然(ML)估计从训练数据中获得图象各区域的先验概率分布,采用Bayes方法,在观测数据基础上,根据分割图象的后验分布所对应的MRF模型的条件概率,利用Metroplis采样器获得最大后验概率(MAP)准则下的图象分割。通过对MSTAR的样本目标图象应用该算法,结果表明它可以获得稳健和准确的分割效果。

References

[1]  [2]Cook R, Mcconnell I. Mum segmentation for SAR images[A].In:SPIE Proc. [C]. Rome, 1994,2316:92~103.
[2]  [5]Derin H, Elliot H. Modeling and segmenttion of noisy and textured images using Gibbs rabdom field[J]. IEEE Trans. on Pattern Analysis and Machine Intelligence, 1987,9(1):39~55.
[3]  [9]Besag J. On the statistical analysis of dirty pictures[J]. Journal of Royal Statistical Society. B, 1986,48:259~302.
[4]  [10]Park J, Kurz L. Image enhancement using the modified ICM method [J]. IEEE Trans. on Image Processing, 1996, 5(5):765~771.
[5]  [1]White R G. Change detection in SAR imagery[J]. Int. J. Remote sensing, 1991,12: 339~ 360.
[6]  [3]Geman S, Geman D. Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of images [J]. IEEE Trans. on Pattern Analysis and Machine Intelligence, 1984, 6 (6):721~741.
[7]  [4]Gerhard Winkler. Image analysis, Random fields and Dynamuic Monte Carlo methods[D]. Berlin :Springer-Verlag, 1999.
[8]  [6]Weisenseel Robert A, Karl W Clem MRF-based algorithm for segmentation of SAR images[A]. In:Proceeding of Int. Conf. on Image Processing [C], Chicago, 1998,3: 770~ 774.
[9]  [7]Ulaby F T, Kouyate F. Texture information in SAR images[J].IEEE Trans. on Geosc. Remote Sensing, 1986, 24(2): 235~245.
[10]  [8]Lakshmanan S, Derin H. Simultaneous parameter estimation and segmentation of Gibbs random fields using simulated annealing [J].IEEE Trans. on Pattern Analysis and Machine Intelligence, 1989,11(8) :799~813.

Full-Text

Contact Us

service@oalib.com

QQ:3279437679

WhatsApp +8615387084133