Grain segmentation in sandstone thin-section based on computer analysis of microscopic images

Abstract The segmentation of minerals visible in sandstone thin-sections is a necessary part microscopic estimation mineral composition, the most important petrographic and sedimentological studies. This process often hard task due to difficulty determining exact boundaries between grains, mainly caused by secondary alteration minerals, etching grains pore-space filling cements. Structural features play very role identification undoubtedly without their use recognition thin sections gives many misclassification results. Calculation each grain area an that time-consuming if done by-hand. Presented method provides precise solution while mixing automated non-automated approach. Photos section were taken using Nikon Eclipse LV100N POL polarizing microscope, at 200x magnification, transmitted light, with crossed polarizers. Then determine borders therefore distinct sample image has been chosen. Initial contours have created by-hand inside graphic application on tablet device, as presented method. Afterward, layer marked was analyzed computer software. In method, detected algorithm areas calculated afterward.