رضا دیانت

This paper represents a tow-step approach to improve de-speckling in SAR images. Firstly, a de-noising method based on modified Bayesian thresholding is applied on decomposed wavelet coefficients. This estimator is based on maximum a posterior which against classical one, can solve complexity and time consuming problems. All traditional speckle reduction approaches cause artificial structures, blurred and smoothed image, so the learning of a Kohonen self organizing map (SOM) is performed directly on the de-noised image to take out the blur. Quantitative and qualitative comparisons of the results obtained by the new method with the results achieved from the other speckle noise reduction techniques demonstrate its higher performance for speckle reduction in SAR images.