SAR Image Modeling Based on the Mixture Gamma Heavy-tailed Rayleigh Distribution

Abstract

Abstract:

Aimed at normal resolution SAR image can't model accurately for the high resolution SAR image，this paper proposes an efficient statistical model, called mixture Gamma heavy-tailed Rayleigh distribution specially for high resolution SAR image. In this model，we model the speckle using the Heavy-tailed Rayleigh density to describe the characteristic of sharp peak and heavy tail of the high resolution SAR image. And we model the RCS by introducing the mixture Gamma density to characterize the complex fluctuation characteristics of the high resolution SAR. Based on the Mellin transformation，we derive mixture Gamma heavy Rayleigh distribution log- cumulant Parameter estimation formula to improve the accuracy of parameter estimation, therefore, to achieve accurate modeling of the high resolution SAR image. Furthermore, experimental results on several actual SAR images are given and we compare our method with some existing methods. Experimental results demonstrate that the modeling proposed in this paper can model for various high resolution SAR images，and it has higher fitting accuracy.

Key words: SAR image, multiplicative model, heavy-tailed Rayleigh distribution, mixture Gamma distribution, Mellin transform

CLC Number:

TP751.1

References

参考文献
[1] HUANG Shiqi, LIU Daizhi. Research on Method and Application of Speckle Noise Reduction of SAR Image[J].Acta Geodaetica et Cartogrphica Sinica, 2006, 35(3): 245-250.(黄世奇，刘代志. SAR 图像斑点噪声抑制方法与应用研究[J]. 测绘学报. 2006, 35(3): 245-250.)
[2] Soumekh M. Synthetic aperture radar signal processing[M]. Wiley New York, 1999.
[3] WU Yan, JIAO Jingmei, YANG Xiaoli. Segmentation Algorithm for SAR Images Based on Fusion of HMT in the Contourlet Domain and D -S Theory of Evidence[J].Acta Geodaetica et Cartogrphica Sinica,2011, 40(2): 148-155.(吴艳，焦惊眉，杨晓丽，等. 基于 Contourlet 域 HMT 和 DS 证据融合的 SAR 图像分割[J]. 测绘学报. 2011, 40(2): 148-155.)
[4] ZHANG Zhongshan, YU Jie, YAN Qin. Research on Polarimetric SAR Image Speckle Reduction Using Kernel Independent Component Analysis[J].Acta Geodaetica et Cartogrphica Sinica,2011, 40(3): 289-295.(张中山，余洁，燕琴，等. 基于核独立成分分析的极化 SAR 图像相干斑抑制[J]. 测绘学报. 2011, 40(3): 289-295.)
[5] WAN Honglin, JIAO Licheng, WANG Guiting. A Region-of-interest Level Method for Change Detection in SAR Imagery[J].Acta Geodaetica et Cartogrphica Sinica,2012, 41(002): 239-245.(万红林，焦李成，王桂婷，等. 在感兴趣的区域层面上进行 SAR 图像变化检测的方法研究[J]. 测绘学报. 2012, 41(002): 239-245.)
[6] Arsenault H H, April G. Properties of speckle integrated with a finite aperture and logarithmically transformed[J]. JOSA. 1976, 66(11): 1160-1163.
[7] Ward K D. Compound representation of high resolution sea clutter[J]. Electronics letters. 1981, 17(16): 561-563.
[8] Kuruoglu E E, Zerubia J. Modeling SAR images with a generalization of the Rayleigh distribution[J]. Image Processing, IEEE Transactions on. 2004, 13(4): 527-533.
[9] Achim A, Kuruoglu E E, Zerubia J. SAR image filtering based on the heavy-tailed Rayleigh model[J]. Image Processing, IEEE Transactions on. 2006, 15(9): 2686-2693.
[10] Jao J. Amplitude distribution of composite terrain radar clutter and the K-Distribution[J]. Antennas and Propagation, IEEE Transactions on. 1984, 32(10): 1049-1062.
[11] Li H C, Hong W, Wu Y R, et al. An efficient and flexible statistical model based on generalized Gamma distribution for amplitude SAR images[J]. Geoscience and Remote Sensing, IEEE Transactions on. 2010, 48(6): 2711-2722.
[12] HUANG Shiqi, LIU Daizhi, HU Mingxing, et al. Multi-temporal SAR Image Change Detect ion Technique Based on Wavelet Transform[J].Acta Geodaetica et Cartogrphica Sinica,2010, 39(2): 180-186.(黄世奇，刘代志，胡明星，等. 基于小波变换的多时相 SAR 图像变化检测技术[J]. 测绘学报. 2010, 39(2): 180-186.)
[13] Achim A, Tsakalides P, Bezerianos A. SAR image denoising via Bayesian wavelet shrinkage based on heavy-tailed modeling[J]. Geoscience and Remote Sensing, IEEE Transactions on. 2003, 41(8): 1773-1784.
[14] Tison C, Nicolas J M, Tupin F, et al. A new statistical model for Markovian classification of urban areas in high-resolution SAR images[J]. Geoscience and Remote Sensing, IEEE Transactions on. 2004, 42(10): 2046-2057.
[15] SHI Ggongtao, ZHAO Lingjun, GUI Lin, et al.A Novel Parameter Estimaton Method for the K Distribution Based on the Mellin Transform[J].Acta Electronica Sinica,2010, 38(009): 2083-2089.(时公涛，赵凌君，桂琳，等. 基于 Mellin 变换的 K 分布参数估计新方法[J]. 电子学报. 2010, 38(009): 2083-2089.)
[16] SUN Zengguo，HAN Chongzhao. Modeling high-resolution synthetic aperture radar images with heavy-tailed distributions[J].Acta Physica sinica,2010,59(02)998-1008.(孙增国，韩崇昭. 基于拖尾分布的高分辨率合成孔径雷达图像建模[J]. 物理学报. 2010, 59(02)998-1008.)
[17] Oliver C, Quegan S. Understanding synthetic aperture radar images[M]. SciTech Publishing, 2004.
[18] NI Kuangyao, GAO Gui,et al. Synthetic Aperture Radar Target Detection Theory Algorithms and Applications[M]. National University of Defense Technology Press,2007.(匡纲要, 高贵，蒋咏梅等. 合成孔径雷达目标检测理论, 算法及应用[M]. 国防科技大学出版社, 2007.)
[19] Nicolas J M. Introduction to second kind statistics: application of Log-moments and Log-cumulants to SAR image law analysis[J]. Traitement du Signal. 2002, 19(3): 139-168.
[20] Delignon Y, Pieczynski W. Modeling non-Rayleigh speckle distribution in SAR images[J]. Geoscience and Remote Sensing, IEEE Transactions on. 2002, 40(6): 1430-1435.