This repository contains Supporting Material to accompany Chapter 5 of my PhD Thesis submitted to the University of Bristol

Below, you'll find a brief description of directories in this repository

CNN_VolumeSegmentation

• Python PyTorch code for training Convolutional Neural Networks to detect endolithic features in coral µCT X-ray slices

RCode

• R code used to produce figures

Data

• Spreadsheets containing data for age and stress event analyses
• Links to the training datasets used and trained models
• Histograms Contains all 16-bit histograms of segmented endolithic features

Utils

• Python code for mass import of groundtruth µCT X-ray slices onto TagLab for easy labelling
• Python code to produce augmented CNN training datasets
• Python code for densitometry
• Python code for manipulation of X-ray label slices and filename fixing