Efficient BFCN for Automatic Retinal Vessel Segmentation

Figure 1

The dataset used. The first row is the original image, and the second row is the ground truth. (a) DRIVE, (b) STARE, and (c) CHASE.

(a)

(b)

(c)