On the space-varying filtering

Filtering of two-dimensional noisy signals is considered in the paper. The concept of nonstationary space-varying filtering is used. It has been shown that the space-varying filtering, along with the algorithm for adaptive Wigner distribution (WD) estimation, may be efficiently used in filtering of two-dimensional signals with high amount of noise, when the noise overlap with signal in space and frequency separately, but not significantly in the joint space/spatial-frequency domain.