Skip to main content
. 2020 Aug 3;15(8):e0236493. doi: 10.1371/journal.pone.0236493

Fig 8.

Fig 8

Segmentation results for existing methods and the proposed method: (a) original input image, (b) ground-truth segmentation map, and segmentation maps generated by (c) SegNet, (d) U-net, and (e) Proposed method.