HMT与HMRF联合的SAR图像小波去斑方法
doi: 10.3724/SP.J.1146.2007.00928
Wavelet-Based Speckil Filtering Comebining HMT and HMRF for SAR Images
-
摘要: 该文将隐马尔可夫树(HMT)和隐马尔可夫随机场(HMRF)两种模型相结合，提出了一种新的估计SAR图像小波系数隐状态的迭代算法。使用该算法可以充分利用小波系数尺度间和尺度内的相关性，更准确地估计隐状态。在此基础上，通过贝叶斯估计分离出小波系数中的信号成分即可消除噪声影响。实验结果表明，该算法能够有效抑制SAR图像相干斑，同时可较好地保持边缘等图像结构特征。Abstract: A novel iterative algorithm to estimate the hidden states of wavelet coefficients of SAR images is proposed by combing the Hidden Markov Tree (HMT) and the Hidden Markov Random Field (HMRF) model. Inter-scale and intra-scale correlation of the coefficients are utilized efficiently in the new approach so the estimation for states is more accurate. With states known the Bayesian estimation is applied to the coefficients to eliminate noises infection. Experiments show that not only the image is despeckled efficiently but also the details are preserved well.
-
[1] 贾承丽, 匡纲要. SAR图像去斑方法. 中国图象图形学报, 2005,10(2): 135-141. [2] Hua Xie, Pierce L E, and Ulaby F T. SAR speckle reductionusing wavelet denoising and Markov random field modeling[J].IEEE Trans. on Geoscience and Remote Sensing.2002,40(10):2196-2212 [3] Crouse M S, Nowak R D, and Baraniuk R G. Wavelet-basedstatistical signal processing using hidden Markov models[J].IEEE Trans. on Signal Processing.1998, 46(4):886-902 [4] Arsenault H H and April G. Properties of speckle integratedwith a finite aperture and logarithmically transformed. J. Opt.Soc. Am, 1976, 11(22): 1160-1168. [5] Donoho D L. Adapting to unknown smoothness via waveletshrinkage[J].J. Amer. Stat. Assoc.1995, 90(432):1200-1224 [6] Tomasi C. Estimating gussian mixture densities with EMa tuorial. www.cs.duke.edu/courses/spring04/cps196.1/handouts/ EM/tomasiEM. pdf, 2004, 3. [7] Besag J. On the statistical analysis of dirty pictures. J. R.Statist. Soc. B, 1986, 48(3): 259-302. -
计量
- 文章访问数: 3707
- HTML全文浏览量: 95
- PDF下载量: 868
- 被引次数: 0
下载:
