Digitization of Petrographic Data for Standardization and Genetic Interpretation

The article describes the methodology of digital analysis of terrigenous rocks from petrographic images obtained with the help of an optical microscope. The developed methodology is aimed primarily for automatic analysis of the structure, composition of rock-forming components and pore space, but also allows evaluating the type and composition of cement, autigenic minerals. The result of proposed analysis is a standardized set of quantitative characteristics, which includes the linear dimensions of each grain and distribution parameters (kurtosis, asymmetry, dispersion, sorting, modal, median and maximum values), shape characteristics (area, roundness), grain - texture packing parameters of the sample (linear and area density, contact, morphology of contacts, orientation of minerals, uniformity of packaging, uniformity of distribution). The set of characteristics also includes information about the pore space: the total pore area, hydraulic coefficient, average pore size and connectivity, the shares of intra-and intergranular pore space are additionally calculated. In addition, the set of quantitative characteristics is supplemented by the result of classification of individual rock-forming minerals by composition and secondary changes, and the classification by mineral composition is carried out for autigenic minerals. The methodology and the result of calculation of these characteristics in this work are demonstrated by the example of a test sample (from the reservoir rock). Petrographic analysis was carried out for this sample in advance and digital pictures were taken with the help of an optical microscope. Automation and digitalization of such analysis became possible only in conditions when the following problems are solved: computer vision problem (segmentation of individual minerals, classification of their composition and secondary changes based on images taken under an optical microscope), computational geometry problem (calculation of numerical characteristics given above, based on binarized images of “masks" of individual pores and minerals).