This paper presents a new adaptive filtering method for multi-temporal Pol(In)SAR images' denoising. Based on the polarimetric scattering mechanism of the given PolSAR data, the filtering first classifies all pixels into persistent scatterer pixels (PS pixels) or distributed scatterer pixels (DS pixels). Then they are separately processed according to its nature. The proposed filtering method has been tested on 31 full polarization Radarsat-2 SAR images of Barcelona Airport area. The results show that the speckle has been well reduced while details have been preserved, which indicates the effectiveness of this filtering approach. © VDE VERLAG GMBH  Berlin  Offenbach.