10.24433/CO.5016803.V1
Weilin Fu
Friedrich-Alexander-Universität Erlangen-Nürnberg
Frangi-Net on High-Resolution Fundus (HRF) image database
Code Ocean
2019
Capsule
Capsule
Computer Science
known-operator-learning
deep-learning
2019-04-24
en-US
3266c510-9558-4997-9201-332c6506be74
5012578
http://dx.doi.org/10.1007/978-3-662-56537-7_87
10.1007/978-3-662-56537-7_87
1.0
GNU General Public License (GPL)
No Rights Reserved (CC0)
This capsule holds the code for Frangi-Net experiment on High-Resolution Fundus (HRF) Image Database.
We reformulate the conventional multi-scale 2-D Frangi vesselness measure into a pre-weighted neural network ("Frangi-Net"). Without training, Frangi-Net is equivalent to the original Frangi filter. With further training, the segmentation performance of the network is increased.