Investigation of the relations between the physical-mechanical and abrasive properties of rocks and the texture coefficient and the calculation of the rock texture coefficient using image processing technique

Rocks are complex heterogeneous masses. The mechanical properties of a rock vary depending on its mineralogy, texture, structure and weathering. In the first stage of this study, 20 rock samples of different formations (magmatic, metamorphic and sedimentary) obtained from different regions were examined for physical-mechanical (Schmidt, porosity-density, Brazilian strength, uniaxial compressive strength, point loading strength) and corrosivity (Los Angeles abrasion resistance, micro deval abrasion resistance and Böhme abrasion resistance) properties were determined. XRF analyzes of the rocks were performed and thin sections were taken from these rocks, thin section images were photographed and analyzed with a polarizing microscope. Mineralogical and petrographic identification was made using XRF and thin section images. In the second stage of the thesis study, the texture coefficient value, which was calculated by visually examining the grains through thin sections with traditional methods and estimating the necessary values, was calculated by developing an original software that calculates it much faster and more accurately using segmentation and deep learning-based image processing techniques. Considering the constant updating of libraries, the multiplicity and functionality of image processing methods, and the applicability criteria in order to calculate the texture coefficient with image processing methodologies, it is predicted that the Python programming language is more useful than other programming languages. For this reason, Python software language was used to calculate the "texture coefficient" using the image processing method. Segment Anything algorithm was used in the segmentation of each mineral grain that forms the texture. Thanks to the developed software, texture coefficient (DK) values of the rocks included in the study were determined.