Edge-preserving noise filtering based on adaptive windowing

An adaptation procedure is introduced for determining in real-time the extent on the analysis window in point estimation of signals corrupted by additive noise. In the tasks of restoring a noisy one-dimensional test signal and a two-dimensional noisy image, mean, median and minimum-mean-square-error filters are compared with fixed and adaptive window implementations. The visual results of the signal and image restorations exhibit superior preservation of edge and detail and suppression of noise for the filters with adaptive windows.<
>