Color image interpolation filter based on adaptive noise detection

In this paper, we present a novel color image interpolation filter for removing salt & pepper impulse noise. Maximum-minimum noise detector and adaptive corrupted degree method are combined to identify pixels which are likely to be corrupted by noise. Then the image is restored by an iterated interpolation estimation method which repairs each noise pixel using its neighbors. Simulation is done for all various noise densities from 10% to 40%, and the results show that the proposed filter can restrain noise with better detail-preserving.