THE NORMALIZED OPTIMIZED ANISOTROPIC WAVELET COEFFICIENT (NOAWC) METHOD - AN IMAGE-PROCESSING TOOL FOR MULTISCALE ANALYSIS OF ROCK FABRIC

The 2D Anisotropic Wavelet Transform (2DAWT) is an image analysis tool which is able to decipher signals where information obtained from dif ferent scales are intermixed. Extended from the Optimised Anisotropic Wavelet Coefficient method (OAWC) of Ouillon et al. [1995], we present a method which discriminates the objects and groups of objects depend ing on their area, shape ratio, orientation and position. Illustrated in a synthetic example, we show that this method allows one to disting uish between different sub-populations of objects within a single phas e, and quantify the anisotropies of shape, orientation and spatial dis tribution at different scales (objects, clusters of objects, alignment s of objects or clusters). Applied to a natural rock sample (Sidobre g ranite, Montagne Noire, France), the 2DAWT has permitted us to detect and accurately characterise the different levels of mineral organisati on, and thus, to contribute to the understanding of the physical proce sses, such as crystallisation, fluid migration, deformation, etc... re sponsible for such organisations.