期刊文献+

区域Gamma混合模型的SAR图像分割 被引量:7

Synthetic aperture radar image segmentation using a regional gamma mixture model
原文传递
导出
摘要 针对传统Gamma混合模型用于SAR图像分割时忽略像素间空间相关性,导致分割结果不连续并产生大量误分割的现象,提出了区域Gamma混合模型的SAR图像分割算法。首先对图像进行分水岭分割,得到过分割区域块,然后将其作为输入样本进行基于Gamma混合模型的聚类,在模型的参数估计过程中进一步考虑区域间的空间相关性,设计邻域因子融入到迭代过程,得到邻域加权类分布概率。该算法充分利用像素间的空间相关性,能够降低噪声对分割结果的影响。通过合成图像和真实SAR图像的实验表明,本文算法能够实现SAR图像的准确分割。 The application of gamma mixture model on Synthetic Aperture Radar (SAR) image segmentation ignores spatial correlation between pixels. To solve this problem, this study proposes a new algorithm using regional gamma mixture model, which improves segmentation accuracy and reduces speckle noise effects on SAR images. The special imaging mechanism of SAR images results in severe distractions to speckle noise. The traditional gamma mixture model for SAR image segmentation only uses the gray information of images and ignores the spatial correlation between pixels, leading to segmentation result discontinuity and false segmentation. Watershed segmentation is a common initial segmentation algorithm. The defect of this algorithm is the over-segmentation phenomenon, which becomes more serious as the noise increases. Segmentation results cannot accurately represent the ground object features. Therefore, related algorithms should be used to combine segmentation results. A substantial amount of research has been conducted to solve the above problems. This study proposes a new segmentation algorithm that increases both segmentation accuracy and noise immunity. This algorithm involves two main steps: building of a regional gamma mixture model and designing of a neighborhood factor. SAR images are segmented by the watershed algorithm. The over-segmented results are used as initial clustering objects for the gamma mixture model. However, the watershed algorithm has a serious problem of oversegmentation. The watershed algorithm and gamma mixture model are combined to include regional information into calculation; however, inter-regional linkages remain missing, and segmentation accuracy remains poor. Therefore, considering the correlation between regional blocks in the parameter estimation, we include the neighborhood factor into the expectation-maximization iterative algorithm. The neighborhood factor is the weighted probability of each region belongs to each segmentation class. After several iterations, the factor is updated, and the image is segmented. Synthetic images and real SAR images demonstrate that the new algorithm can achieve more accurate segmentation results than the traditional gamma mixture model, regional Gaussian mixture model and regional Markov random field algorithm. Upon visual observation, the segmentation results of the new algorithm are clearer and less affected by noise. Moreover, the results of the new algorithm have larger Kappa coefficient and better segmentation accuracy. This study proposes the RGaMM algorithm, which includes regional information and interregional contex- tual information into calculation. This algorithm can also effectively segment SAR image. In particular, under the effects of heavy speckle noise, the RGaMM algorithm has more advantages than the traditional algorithm. The proposed algorithm not only can achieve accurate segmentation on SAR images, but also reduce speckle noise effects.
出处 《遥感学报》 EI CSCD 北大核心 2014年第5期1024-1033,共10页 NATIONAL REMOTE SENSING BULLETIN
基金 国家自然科学基金(编号:61371154 41076120 61271381 61102154) 光电控制技术重点实验室和航空科学基金联合资助项目(编号:201301P4007) 中央高校基本科研业务费专项基金资助(编号:2012HGCX0001)
关键词 合成孔径雷达 分水岭 Gamma混合模型 邻域因子 EM算法 synthetic aperture radar, watershed, gamma mixture model, neighborhood factor, expectation-maximization algorithm
  • 相关文献

参考文献17

  • 1Ben Ayed I, Mitiche A and Belhadj Z. 2005. Multiregion level-set parti- tioning of synthetic aperture radar images. IEEE Transactions on Pattern Analysis and Machine Intelligence, 27(5): 793-800 [DOI: 10.1109/TPAMI.2005.106 ].
  • 2Celik T and Lee H K. 2013. Comments on "A robust fuzzy local infor- mation C-means clustering algorithm". IEEE Transactions on Images Processing, 22 (3): 1258 - 1261 [DOI: 10.1109/TIP.2012. 2226048 ].
  • 3Dabboor M, Collins M J, Karathanassi V and Braun A. 2013. An unsu- pervised classification approach for polarirnetdc SAR data based on the Chemoff distance for complex Wishart distribution. IEEE Trans- actions on Geoscience and Remote Sensing, 51 (7): 4200-4213 [ DOI:10.1109/TGRS20122227755].
  • 4Deng H W and Clausi D A. 2005. Unsupervised segmentation of synthetic aperture radar sea ice imagery using a novel markov ran- dom field model. IEEE Transactions on Geoscience and Remote Sensing, 43(3): 528-538 [ DOI:10.1109/TGRS2004.839589.
  • 5Doulgeris A P, Anfmsen S N and Eltott T. 2011. Automated non- Gaussian clustering of polarimetric synthetic aperture radar images. IEEE Transactions on Geoscience and Remote Sensing, 49(10): 3665-3676 [ DOI:10.1109/TGRS20112140120 ].
  • 6Goodman J W. 1975. Chapter I-Statistical Properties of Laser Speckle Patterns//Laser speckle and related phenomena. New York: Spring- Verlag: 9-75.
  • 7Gou S P, Zhuang X, Zhu H M and Yu T T. 2013. Parallel sparse spectral clustering for SAR image segmentation. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 6 (4): 1949-1963 [DOI:10.1109/JSTARS.2012.2230435 ].
  • 8Haverkamp D, Soh L K and Tsatsoulis C. 1993. A dynamic local thresh- olding technique for sea ice classification//IEEE International Con- ference on Geoscience and Remote Sensing Symposium . Tokyo: IEEE: 638-640 [ DOI:10.1109/IGARSS.1993 322252 ].
  • 9Li Y, Li J and Chapman M A. 2010. Segmentation of SAR intensity im- agery with a voronoi tessellation, bayesian inference, and reversiblejump MCMC algorithm. IEEE Transactions on Geoscience and Remote Sensing, 48 (4): 1872 - 1881 [DOI: 10.1109/TGRS.2009. 2033588].
  • 10卢洁,杨学志,郎文辉,左美霞,徐勇.区域GMM聚类的SAR图像分割[J].中国图象图形学报,2011,16(11):2088-2094. 被引量:16

二级参考文献38

  • 1万成凯,袁保宗,苗振江.一种基于活动轮廓和Gauss背景模型的固定摄像机运动目标分割算法[J].中国科学(F辑:信息科学),2009,39(4):391-396. 被引量:6
  • 2宋余庆,谢从华,朱玉全,李存华,陈健美,王立军.基于近似密度函数的医学图像聚类分析研究[J].计算机研究与发展,2006,43(11):1947-1952. 被引量:16
  • 3Oliver C, Quegan S. Understanding Synthetic Aperture Radar Image[ M]. Boston London:Assech House, 1998.
  • 4Otsu N. A threshold selection method from gray-Level histograms [J ]. IEEE Transactions on Systems, Man and Cybernetics, 1979, 9( 1 ) :62-66.
  • 5Coleman G B, Andrews H C. Image segmentation by clustering[J]. Proceedings of the IEEE, 1979, 67(5) :773-785.
  • 6Adams R, Bischof L. Seeded region growing [ J ]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1994, 16(6) :641-647.
  • 7Bovik A C. On detecting edges in speckle imagery [ J ]. IEEE Transactions on Acoustics, Speech and Signal Processing, 1988, 36(10) :1618-1627.
  • 8Wang Xiaofeng, Zhang Xiaoping. A new localized superpixel Markov random field for image segmentation [ C ]//2009 IEEE International Conference on Multimedia and Expo. New York: IEEE Computer Society,2009 : 642-645.
  • 9Fraley C, Raftery A E. How many clusters .9 Which clustering method .9 Answers via model-based cluster analysis [ J ]. The Computer Journal, 1998, 41 (8) :578-588.
  • 10Vincent L, SoiUe P. Watersheds in digital spaces: an efficient algorithm based on immersion simulations [ J ]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1991, 13(6) :583-598.

共引文献89

同被引文献102

  • 1崔斌,张永红,闫利,魏钜杰.高分三号SAR影像双阈值变化检测[J].遥感学报,2020,24(1):1-10. 被引量:6
  • 2王文光,王俊,毛士艺.一种基于差异度的极化SAR图像迭代分类方法[J].电子与信息学报,2006,28(11):2007-2010. 被引量:9
  • 3侯一民,郭雷.一种基于马尔可夫随机场的SAR图像分割新方法[J].电子与信息学报,2007,29(5):1069-1072. 被引量:28
  • 4Cao F, Hong W, Wu Y R and Pottier E. 2007. An unsupervised segmentation with an adaptive number of clusters using the SPAN/ H/c~/A space and the complex wishart clustering for fully polafime- tric SAR data analysis. IEEE Transactions on Geoscience and Remote Sensing, 45 ( 11 ) : 3454 - 3467 [DOI: 10. ll09/TGRS. 2007. 907601 ].
  • 5Ersahin K, Cumming I G and Ward R K. 2010. Segmentation and classi- fication of polar.metric SAR data using spectral graph partitioning. IEEE Transactions on Geoscience and Remote Sensing, 48 (1) : 164- 174 [ DOI : 10.1109/TGRS. 2009. 2024303 ].
  • 6Lee J S, Grunes M R, Ainsworth T L, Du L J, Schuler D L and Cloude S R. 1999. Unsupervised classification using polarimetric decompo- sition and the complex wishart classifier. IEEE Transaetions on Geoscience and Remote Sensing, 37 (5) : 2249 - 2258 [ DOI: !0. 1109/36. 789621 ].
  • 7Lee J S, Grunes M R and Pottier E. 2001. Quantitative comparison of classification capability: fully polarimetrie versus dual and single- polarization SAR. IEEE Transactions on Geoscience and Remote Sensing, 39( 11 ) : 343 -2351 [DOI: 10. 1109/36. 964970].
  • 8Lee J S and Pottier E. 2009. Polarimetric Radar Imaging : From Basics to Applications. Boca Raton: The Chemical Rubber Company Press: 26 - 43.
  • 9Li S Z. 2009. Markov Random Field Modeling in Image Analysis (3rd edition). New York: Springer-Verlag: 69 -71.
  • 10Pottier E and Lee J S. 1999. Application of the H/A/ ct polarimetric decomposition theorem for unsupervised classification of fully polari- metric SAR data based on the wishart distribution// Proceeding of the Committee on Earth Observing Satellites SAR Workshop, Tou- louse, France: NASA: 335 - 340.

引证文献7

二级引证文献33

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部