全极化数据的可得,使得利用极化散射矩阵以减少多视全极化SAR图像中的相干斑成为了可能。其中,被称为最优的多视极化白化滤波器（MPWF）就是一个典型例子。然而,为提高滤波器参数估计的精度,同时能够自适应的检测图像中的结构特征,在滤波时采用适当的窗算法是必要的。本文提出了两个有效的窗算法,并用1994年NASA SIR-C/X-SAR,L band,经过四视处理的天山森林的极化数据进行了仿真。实验结果表明,这两种算法不仅很好的达到了降噪效果,而且对图像的纹理信息以及结构特征（即边缘特征、线性特征、强散射体特征）的保持效果是最佳的。

The availability of full polarimetric SAR data made it become possible to use three complex elements(HH,HV,VV) of the polarimetric scattering matrix to reduce speckle in multilook full polarimetric SAR images.As a typical example,the multilook polarimetric whitening filter(MPWF) is regarded as optimal.However,to enhance the precision of parameter estimation and meanwhile detect the structural feature adaptively,windowing algorithm must be needed when filtering.In this paper,two new windowing approaches are introduced and NASA SIR-C/X-SAR,L band,four-look processed polarimetric SAR data of the Tian-Mountain Forest is used for simulation.Experimental results demonstrate the effectiveness of these methods both on speckle reduction and preservation of texture information.