Skip to main content
. 2021 Apr 14;16(4):e0247388. doi: 10.1371/journal.pone.0247388

Fig 5. Samples of the retinal fundus images and their H×W dimensions in pixels.

Fig 5

(a) Input images: Left: DRIVE (584×565); right: CHASE_DB1 (960×999). (b) Manual annotation of the retinal vessels. (c) Binary segmentation masks of the fundus images.