Date Added: Dec 2009
In this paper a novel method for effectively denoising the extremely corrupted image by fixed value impulse noise using robust estimation based filter is proposed. The proposed algorithm classifies the pixels of localized window in to "Corrupted" or "Uncorrupted" and removes only corrupted pixels by robust estimation or by mean of the processed neighboring pixels. It is shown that the proposed filter effectively removes the impulse noise while preserving the good image quality. The visual and quantitative results prove that the performance of proposed filter in the preservation of edges and details is better even at noise level as high as 95%.