A Computer Aided System for Reservoir Rock Classification

Abstract

This study is mainly aimed to know the reservoir rocks among core specimens that are obtained during the stage of drilling of exploration oil and gas wells by the processing of the image digitally using a Computer-Aided System. This computerized method has been conducted using two programming languages; Python and Matlab. The Python program is used to segment rock images and classify the extracted features, whereas the MATLAB language has been used to extract intensity features from segmented images. The database used in this project has been collected from two sources, one of them from the Sarier formation (the biggest Libyan oil field) while the second source was from the internet. The images used in this study are sedimentary rocks images, which include sandstone, limestone, and dolomite. As a result, this study has reached that the system, which was used here is capable of classifying rocks and can be used to classify oil rocks or any other type of rocks. Also, it can be used near the well due to certain conditions such as the absence of experts due to the lack of enough database, the study has only introduced the preliminary results as proof of the concept.