A Novel Statistical Approach for Speckle Filtering of SAR Images

Synthetic aperture radar (SAR) images are inherently affected by a signal multiplicative noise known as speckle, which is due to the radar wave coherence. In this paper, we introduce a new method to reduce speckle noise in SAR images, based on statistical modeling of wavelet coefficients. We use two-dimensional generalized autoregressive conditional heteroscedasticity (GARCH) model for statistical modeling of images wavelet coefficients. By using this model, we are capable of taking into account important characteristics of wavelet coefficients, such as heavy tailed marginal distribution, and the dependencies between the coefficients. Furthermore, we use maximum a-posteriori (MAP) estimator for estimating the clean image wavelet coefficients. Finally, we compare our proposed method with some denoising methods, and quantify the achieved performance improvement.