SERR-U-Net: Squeeze-and-Excitation Residual and Recurrent Block-Based U-Net for Automatic Vessel Segmentation in Retinal Image

<div>Segmentation results on DRIVE and STARE: (a, f) original retinal image; (b, g) grayscale conversion; (c, f) binarization; (d, i) our result; (e, j) expert 2<sup>nd</sup>.</div>

Computational and Mathematical Methods in Medicine

fig9

Figure 9

Figure 9: SERR-U-Net: Squeeze-and-Excitation Residual and Recurrent Block-Based U-Net for Automatic Vessel Segmentation in Retinal Image 