Automatic characterization of texture of carbonate rocks from integrated analysis of micro-resistivity borehole images and petrographic image analysis

	Information on texture is extracted from electrical images and associated to rock facies using an innovative (and patented) approach combining wavelet analysis, neural network self-organising classification scheme and ground-truth data obtained from quantitative petrographic image analysis of exhaustive thin-section study.
	Grain-supported facies (double porosity) Mud-supported facies (single porosity)

The interpretation of RMN downhole measurements accounts for the rock facies to allow correct determination of the porosity (Illustrations courtesy of Schlumberger Ltd, taken from Ramakrishnan, Rabaute et al. (1998) and Saito, Rabaute, Ramakrishnan (2001).) The project was funded by Schlumberger-Doll Research Lab.