%A NIM. 12650030 MUHAMMAD DZULFIKAR FAUZI
%O Dr. Shofwatul ‘Uyun, S.T., M.Kom
%T PENGOLAHAN CITRA SAYATAN TIPIS MENGGUNAKAN ALGORITMA MULTILEVEL OTSU’S THRESHOLDING PADA BATUAN BEKU
%X The image of a thin section of rock is a rock or mineral observation methods. The mineral content of rocks can be used to determine rock types and names of these rocks. The main function of a thin section of rock is to do the analysis and identification of minerals in rocks. Thin section making observations difficult if only using polarizing microscopy, the ability of the human eyes can not identify in detail. Therefore, the analysis and identification of minerals to be more easily done with the help of digital image processing algorithms using Multilevel Otsu's thresholding to determine the abundance of igneous minerals.
The data used in this study using the image of a thin section of igneous rock acquired from use polarizing microscope with a magnification of 10x, the next stage perform preprocessing, processing, and analysis. Preprocessing stage used is cropping, k-means with value of k=5,6 and 7, and grayscaling. While the processing stage is segmentation the image. Segmantasi used is Multilevel Otsu's thresholding with two threshold values. Furthermore, the analysis phase containing mineral abundance calculation based on the area of segmentation results.
The results of all the calculations Confusion Matrix that has been done shows that the naming of Mela- Gabbro, Gabbro and Leuco- Gabbro rock have an accuracy rate of 67%, 58% and 0%. In both scenarios have been for naming gabbro rocks using Otsu's Multilevel thresholding algorithm and K-means clustering with k = 6 and 7.
%K Image thin section, Igneous rock, Multilevel Otsu’s Thresholding
%D 2016
%I UIN SUNAN KALIJAGA YOGYAKARTA
%L digilib21306