This repository contains the code and resources for a computer vision project focused on enhancing grain boundary detection and segmentation in quartz rocks. This project is a collaboration between four team members and is aimed at assisting Matty Mookerjee, head of the geology department at SSU. The primary focus is on analyzing EBSD (Electron BackScatter Diffraction) and EDS (Electron Dispersion Spectroscopy) images.
Luis A Carmona Zander Joel Thomas Ben Harris
The project aims to: Develop a program to convert CSV files containing EBSD and EDS data into images for analysis. Research and potentially utilize chemistry transfer learning models to enable effective communication between datasets (linking chemical compositions of rocks with their band contrasts). Experiment with the StarDist library for potential improvements in segmentation processes. Explore various morphological techniques for enhanced image analysis.
We can include steps for installing any required libraries, setting up the environment, and how to run the scripts when the time comes.
We can describe the data we used, its format, and how to access it, if applicable.
We can summarize any key results, findings, or insights we gained from the project.