Directional multi-scaled fusion based median filter for removal of RVIN

Abstract

In this paper Directional Multi-scaled Fusion Based Median Filter for Removal of RVIN (DMFBMF) has been proposed to restore images corrupted with random valued impulse noises. This filter employs a technique of impulsive noise detection-suppression strategy from images, which is simple but efficient and uses a fusion technique to suppress the impulses from the corrupted images. At first, using two methodologies two different restored images are obtained from the same corrupted source image. Then using a simple fusion rule these two restored images are fused. Thus, a well impulse suppressed image has been obtained. Extensive simulations showed that this scheme provides good performances of suppressing impulse with low noise level as well as for highly corrupted benchmarked images.