Caprari R S, Goh A S, Moffatt E K
Defence Science and Technology Organisation, PO Box 1500, Salisbury SA 5108, Australia.
Appl Opt. 2000 Dec 10;39(35):6633-40. doi: 10.1364/ao.39.006633.
We present a Wiener filter that is especially suitable for speckle and noise reduction in multilook synthetic aperture radar (SAR) imagery. The proposed filter is nonparametric, not being based on parametrized analytical models of signal statistics. Instead, the Wiener-Hopf equation is expressed entirely in terms of observed signal statistics, with no reference to the possibly unobservable pure signal and noise. This Wiener filter is simple in concept and implementation, exactly minimum mean-square error, and directly applicable to signal-dependent and multiplicative noise. We demonstrate the filtering of a genuine two-look SAR image and show how a nonnegatively constrained version of the filter substantially reduces ringing.
我们提出了一种维纳滤波器，它特别适用于多视合成孔径雷达（SAR）图像中的斑点和噪声抑制。所提出的滤波器是非参数的，不基于信号统计的参数化分析模型。相反，维纳 - 霍普夫方程完全根据观测到的信号统计来表示，而不涉及可能不可观测的纯信号和噪声。这种维纳滤波器在概念和实现上都很简单，具有精确的最小均方误差，并且直接适用于信号相关和乘性噪声。我们展示了对一幅真实的两视SAR图像的滤波，并展示了该滤波器的非负约束版本如何显著减少振铃现象。
