Image Diffusion Methods with Integral-based Diffusion Coefficient

Diffusion filters are widely used to reduce noise on images. In [1] a new diffusion method for image filtering based on incorporating an integral of image intensity over a point neighborhood into the diffusion coefficient was suggested. Here we start by introducing two modified diffusion filters. Next we illustrate on number of images the advantages of modified filters comparing with not modified ones. In last section we introduce image sharpening method based on backward diffusion problem. We use constant diffusion coefficient and coefficient constructed using the incorporating an integral of image intensity over a point neighborhood. Benefits of using modified filter shown on the set of images.