Learning Fully-Connected CRFs for Blood Vessel Segmentation in Retinal Images

Explore this paper's citation graph

Summary

This work presents a novel method for blood vessel segmentation in fundus images based on a discriminatively trained, fully connected conditional random field model with more expressive potentials, and employs recent results enabling extremely fast inference in a fully connected model.

Type
article
Published
2014-09-14
Cited by
146
References
30
Access
Open access

Keywords

Computer science, Conditional random field, CRFS, Segmentation, Artificial intelligence

References

Cited by

Related papers