A Volterra type model for image processing

Abstract
We present a class of time-delay anisotropic diffusion models for image restoration. These models lead to asymptotic states that are selected on the basis of a contrast parameter and bear some analogy with neural networks with Hebbian dynamical learning rules. Numerical examples show that these models are efficient in removing even high levels of noise, while allowing an accurate tracking of the edges.

This publication has 12 references indexed in Scilit: