Document Scanning in a tough environment : application to cameraphones

Abstract

The aim of this paper is to suggest a new method to enhance text in document-image. In the first step, we present a classical model based on non-convex optimization problem. In this way, a simultaneous estimation of the reflectance and the luminance are obtained when the non uniform illumination (also called luminance) is a smooth function and the reflectance is a function of bounded variation. We prove some conditions of existence and unicity. Secondly, we introduce the ”log” of the classical problem which generate a new PDE’s model. Resolution of this method is based on solving an original Partial Differential Equation (PDE) estimating the log of the luminance. We assume that the luminance is enough smooth and is the solution of a non classical second order’s PDE, and we deduce the reflectance from the estimated luminance and the acquired image. To illustrate the robustness of the proposed process (by means to prove the ability of this method to improve an Optical Character Recognition (OCR) in text recognition) we give a numerical examples in real-world situation (images acquired from cameraphones).