Image Segmentation of Retinal Blood Vessels Based on Dual-Attention Multiscale Feature Fusion

Figure 5

Visualization results of different methods on DRIVE and STARE datasets: (a) original image; (b) ground truth; (c) literature [29]; (d) literature [30]; (e) literature [31]; (f) literature [32]; (g) literature [33]; (h) proposed.

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)