This repository contains code in support of the paper: "Optic Disc and Cup Segmentation Methods for Glaucoma Detection with Modification of U-Net Convolutional Neural Network":
Sevastopolsky, Artem. "Optic Disc and Cup Segmentation Methods for Glaucoma Detection with Modification of U-Net Convolutional Neural Network." arXiv preprint arXiv:1704.00979 (2017).
(accepted for publication in "Pattern Recognition and Image Analysis: Advances in Mathematical Theory and Applications" journal, ISSN 1054-6618)
Built with Python 2.7 and Keras.
See scripts folder for notebooks for training with clarification of usage. HDF5 datasets should be recreated with scripts/Organize datasets.ipynb notebook. models_weights folder contains pre-trained models.
Click the following links to watch content of notebooks in a handy way:
- U-Net, OD on RIM-ONE v3 (fold 0).ipynb (nbviewer)
- U-Net, OD on DRIONS-DB (fold 0).ipynb (nbviewer)
- U-Net, OD cup on RIM-ONE v3, cropped by OD (fold 0).ipynb (nbviewer)
- U-Net, OD cup on DRISHTI-GS, cropped by OD (fold 0).ipynb (nbviewer)
The software is distributed under MIT License, which requires that copyright notice and this permission notice shall be included in all copies or substantial portions of this software. Commercial use, distribution, modification and private use are fully allowed.