Decision Based Adaptive Neighborhood Median Filter

Impulse noise in an image degrades the performances of the image processing and analysis stages. Therefore the noise removal and correction is an important processing required before performing any subsequent image processing approaches in the image data. In this paper a novel approach is proposed to remove the impulse noise from gray images. In this technique, all the pixels having gray value 0 and 255 are considered as the noisy pixels. The proposed method consists of three stages of filtering by considering the neighborhood pixels. The proposed method outperforms the Standard median Filter, Improved Fast Peer Group Filter and Modified Decision Based Unsymmetric Trimmed Median Filter.