Synthetic aperture radar (SAR) images are disturbed by a multiplicative noi
se depending on the signal (the ground reflectivity) due to the radar wave
coherence, Images have a strong variability from one pixel to another reduc
ing essentially the efficiency of the algorithms of detection and classific
ation, In this study, we propose to filter this noise with a multiresolutio
n analysis of the image. The wavelet coefficient of the reflectivity is est
imated with a Bayesian model, maximizing the a posteriori probability densi
ty function. The different probability density function are modeled with th
e Pearson system of distributions, The resulting filter combines the classi
cal adaptive approach with wavelet decomposition where the local variance o
f high-frequency images is used in order to segment and filter wavelet coef
ficients.