Thin Section Mineral Identification and Petrographic analysis

by Jon Barnes for Portland State University

A Python Library of functions that will use machine learning to automate the identification of mineral grains, which grains are similar in composition and which very by trace minerals (for example which grain is biotite or plagioclase or olivine and which ones have more titanium or other elements), and a description of the grain variation in size-composition that will talk about the petrology of the sample (if it has large crystals, porphyritic composition, what percentage is ground mass, etc).