Speckle reduction with edge-preserving

Coherent imagery emerges as one of the major domains in image processing and includes topics as diversified as radar, medical and surface analysis. Whatever the application, the resulting image is noisy corrupted. In coherent imaging, images suffer from speckle noise, whose main characteristic is to be multiplicative. The proposed method takes explicitly into account the multiplicative property of the noise while preserving discontinuities in the restored image. Moreover, in a second step our algorithm estimates the noise, thus the information contained in the speckle still remains usable