Image Processing by Means of a Linear Integro-differential Equation

E. Cuesta and J. Finat Codes (Spain)

Keywords

Image restoration, fractional integrodifferentialequations,fractional quadratures rules.

Abstract

Partial differential equations (PDE) have been consid ered in image processing for denoising and stabilizing edges. Main approaches concern to diffusion processes (heat equation) and variational principles (energy method). In this work we propose a first approach to image process ing by means of integro-differential equations of fractional order. Our purpose is to exploit properties of the solution of the linear fractional integro-differential equation, with special regard to regularization processes. Some of these properties can be considered as intermediate between those of the heat equation and the ones of the wave equation. Practical illustrations are provided using suitable numerical methods, thoses combine both fractional quadrature rules (FQR) (or GrunwaldLetnikov rules), and classical numer ical schemes.

Important Links:



Go Back