This project analyzes images of rock core samples with the following steps:

1. Define areas in the images that contain valid pixels
2. Analyze RGB composition of the valid pixels and perform K-Means clustering
3. Use pre-trained deep learning models (ResNet50, InceptionV3, Xception) for feature extraction and clustering
4. Enhance the edges of the images and apply step 3

The Jupyter Notebook might be too large to be opened on GitHub. If that happens, please find the exported markdown here: Markdown exported from Notebook