A modified GNS map denoising approach for digital images

Abstract

This paper proposes a new technique of noise reduction for 2D digital images. In the proposed model, for constructing a GNS map using DNS maps of central pixels we measured the intensity similarity between center pixel and its neighboring pixels within a certain window. In the proposed scheme, the intensity similarity is being calculated by using Manhattan distance equations; where the conventional GNS map approach uses Euclidean distance principle. To validate the performance of the proposed model, we consider several attacks on our tested dataset and experimental results demonstrate that the proposed model performs better than the conventional GNS map based denoising technique by exhibiting lower MSE results.