A Robust Optimal Morphological Filter to Remove Impulse Noise in Images

Image smoothing is a primary step to remove noise and sharpens the important edge features in the image. The present paper proposes a new Optimal Morphological Filter (OMF) for noise removal in the image. The impulsive noise take gray level values as minimum or maximum from the dynamic gray level range. Optimal notations are proposed in this paper to eliminate the noise pixels corrupted by low and high gray level values in two stages. The method also adopted median and is compared with obtained optimal notation values from local neighborhood of the image to preserve important image features. Different grayscale noise images are tested and visual results show the performance of the proposed algorithm.