Synthetic aperture radar(SAR) images are disturbed by a speckle noise depending on the signal(the ground reflectivity) due to the radar wave coherence. Images have a strong variability from one pixel to another reducing essentially the efficiency of the algorithms of object detection and classification. In this study

we propose to suppress this noise with a multiresolution analysis of the image. Because the Mallat decomposition can not preserve the translation invariance of an image

the filter in this paper is based on stationary wavelet decomposition. This method is used to an ERS-1 SAR image and then it is compared to a filter based on Mallat decomposition and three other classic SAR image filters

the results show that the presented approach has the advantage both in speckle suppression and the preservation of image details.