4 - Analyse multirésolution pour les images avec un facteur de résolution √2

Recently, an algorithm for multiresolution signal analysis, based on tvavelet theory las been proposed by S. Mallat. The basic idea is to capture significant details in the signal through the analysis ai successive scales . Mallat's algorithm uses a scaling factor of 2 between two successive scales . The aim of this paper is to develop a theory and an algorith n for image processing, with a scaling factor of \F2 . This will allow for sinipler resulis interpretation and doubled sharpness of analysis . The theory is illustrated on both artificial and natural images, where our algorithni proves efficient for non-oriented contour detection .