Online supplementary material for paper "Artery-Vein Segmentation in Fundus Images using a Fully Convolutional Network" by Hemelings R., Elen B., Stalmans I., Van Keer K., De Boever P., Blaschko M.B.
Revised March 2023.
This repo contains (1) vizualizations of predictions and ground truth on two public data sets; (2) the evaluation code to facilitate reproducibility; (3) links to the novel AV ground truth for the public HRF image database.
You can download the HRF ground truth and DRIVE evaluation masks at:
The HRF ground truth are directly available from the subdir HFR_AV_GT.
The DRIVE evaluation masks are directly available from the subdir DRIVE_AV_evalmasks.
Please cite as:
@article{hemelings2019avs,
title={Artery-Vein Segmentation in Fundus Images using a Fully Convolutional Network},
author={Hemelings, Ruben and Elen, Bart and Stalmans, Ingeborg and Van Keer, Karel and De Boever, Patrick and Blaschko, Matthew B},
booktitle={Computerized Medical Imaging and Graphics},
year={2019},
doi={https://doi.org/10.1016/j.compmedimag.2019.05.004}
}
