Two intelligent pattern recognition models for automatic identification of textural and pore space characteristics of the carbonate reservoir rocks using thin section images

Abstract

Over the last two decades pattern recognition approaches have attracted engineers to solve real world problems more accurately through the development of computational technology. In the present research, the capabilities of intelligent systems are employed to develop two algorithms for identification of textural and pore space characteristics of carbonate rocks from thin section images. The texture identifier model classifies the images based on Dunham classification, while the porosity analyzer model determines the percentage of each type of pore spaces in the image. The texture identifier model extracts thirteen features to recognize texture type and the porosity analyzer determines percentage of each type of porosity based on eleven features extracting from the thin section image. Finally, two confusion matrixes are used to evaluate the performance of the developed models. The results show that the models perform reliably from the perspective of petroleum geology for studying carbonate reservoir rocks.