evaluating of speckle reduction in sar images based on the hard thresholding of curvelet coefficients

due to damaging effects of speckle noise to information of radar images, reduction of these effects has been considered by many researchers. we discussed speckle reduction of radar images based on curvelet transformation. more specifically we discussed the speckle reduction with emphasizing on preservation of the edges by hard thresholding of curvelet transformation. in this algorithm, first multiplicative speckle noise was transformed into an additive one. the curvelet transform was then taken from logarithmically transformed image and the curvelet coefficients were thresholded by hard thresholding function. the results were compared to the other widely-used adaptive filters. we found that the hard thresholding curvelet based filter offers better results than the other filters. in details, mean square error, normalized mean square error and mean absolute error indices were reduced by 25%, 25% and 14%, respectively. in addition, the curvelet based filter was capable of preserving the edges.