Table 5 Performance evaluation of different models. Significant values are in bold.
Methods | Miou | Precision | Recall | F1-Score |
|---|---|---|---|---|
U-Net | 72.18% | 80.10% | 85.16% | 82.58% |
U-Net++ | 78.66% | 74.96% | 63.36% | 68.65% |
R2U-Net | 72.67% | 58.63% | 69.03% | 63.45% |
Improved U-Net Model 1 | 75.53% | 77.72% | 94.12% | 83.58% |
Improved U-Net Model 2 | 78.33% | 80.76% | 94.46% | 86.31% |
WFU-Unet | 81.68% | 91.43% | 86.63% | 88.93% |
