Nonlinear Diffusion in Graphics Hardware

Abstract

Multiscale methods have proved to be successful tools in image denoising,edge enhancement and shape recovery. They are based on the numericalsolution of a nonlinear diffusion problem where a noisy or damagedimage which has to be smoothed or restorated is considered as initialdata. Here a novel approach is presented which will soon be capableto ensure real time performance of these methods. It is based onan implementation of a corresponding finite element scheme in texturehardware of modern graphics engines. The method regards vectors astextures and represents linear algebra operations as texture processingoperations. Thus, the resulting performance can profit from the superiorbandwidth and the build in parallelism of the graphics hardware.Here the concept of this approach is introduced and perspectivesare outlined picking up the basic Perona Malik model on 2D images.