Components classification of the clastic rock thin-sections based on GIS

The process of manual identification of thin-sections under the polarizing microscope is complex and repetitive. Geographic Information System (GIS) is a system of collecting, storing, managing, computing, analyzing, displaying and describing geospatial information. This paper proposes a way of recognizing and classifying components in clastic rock thin-sections image automatically using the spatial analysis and data management functions of GIS. The different components in clastic rock show different interference colors under orthogonal optical. According to the contrast differences between the components, we can extract the boundary of component and store it in a geodatabase through three steps including noise reduction, image segmentation and unsupervised classification. To deal with differences in size and shape of different component, we can use ISODATA(Iterative Organizing Analysis Technique) and MLC(Maximum Likelihood Classification) functions to classify and store the matrix. This paper provides a convenient tool for identifying, classifying and analyzing the thin-sections.