Table 1. The values of the parameters in the deep learning layers.
| Layers | Input Size | Output Size | Stride Size | Kernel Size |
|---|---|---|---|---|
| 1: Input | 720 × 731 × 1 | 720 × 731 × 1 | — | — |
| 2: Conv | 720 × 731 × 1 | 720 × 731 × 64 | 1 × 1 | 3 × 3 |
| 3: Conv | 720 × 731 × 64 | 720 × 731 × 64 | 1 × 1 | 3 × 3 |
| 4: Conv | 720 × 731 × 64 | 720 × 731 × 64 | 1 × 1 | 3 × 3 |
| 5: Conv | 720 × 731 × 64 | 360 × 366 × 128 | 2 × 2 | 2 × 2 |
| 6: Conv | 360 × 366 × 128 | 360 × 366 × 128 | 1 × 1 | 3 × 3 |
| 7: Conv | 360 × 366 × 128 | 180 × 183 × 256 | 2 × 2 | 2 × 2 |
| 8: Conv | 180 × 183 × 256 | 180 × 183 × 256 | 1 × 1 | 3 × 3 |
| 9: Conv | 180 × 183 × 256 | 90 × 92 × 512 | 2 × 2 | 2 × 2 |
| 10: Conv | 90 × 92 × 512 | 90 × 92 × 512 | 1 × 1 | 3 × 3 |
| 11: Conv | 90 × 92 × 512 | 45 × 46 × 1024 | 2 × 2 | 2 × 2 |
| 12: Conv | 45 × 46 × 1024 | 45 × 46 × 1024 | 1 × 1 | 3 × 3 |
| 13: ConvTranspose | 45 × 46 × 1024 | 90 × 92 × 512 | 2 × 2 | 3 × 3 |
| 14: Conv | 90 × 92 × (512 + 512) | 90 × 92 × 512 | 1 × 1 | 3 × 3 |
| 15: Conv | 90 × 92 × 512 | 90 × 92 × 512 | 1 × 1 | 3 × 3 |
| 16: ConvTranspose | 90 × 92 × 512 | 180 × 183 × 256 | 2 × 2 | 3 × 3 |
| 17: Conv | 180 × 183 × (256 + 256) | 180 × 183 × 256 | 1 × 1 | 3 × 3 |
| 18: Conv | 180 × 183 × 256 | 180 × 183 × 256 | 1 × 1 | 3 × 3 |
| 19: ConvTranspose | 180 × 183 × 256 | 360 × 366 × 128 | 2 × 2 | 3 × 3 |
| 20: Conv | 360 × 366 × (128 + 128) | 360 × 366 × 128 | 1 × 1 | 3 × 3 |
| 21: Conv | 360 × 366 × 128 | 360 × 366 × 128 | 1 × 1 | 3 × 3 |
| 22: ConvTranspose | 360 × 366 × 128 | 720 × 731 × 64 | 2 × 2 | 3 × 3 |
| 23: Conv | 720 × 731 × (64 + 64) | 720 × 731 × 64 | 1 × 1 | 3 × 3 |
| 24: Conv | 720 × 731 × 64 | 720 × 731 × 64 | 1 × 1 | 3 × 3 |
| 25: Conv | 720 × 731 × 64 | 720 × 731 × 1 | 1 × 1 | 3 × 3 |
| 26: Add | 720 × 731 × (1 + 1) | 720 × 731 × 1 | — | — |
| 27: Norm | 720 × 731 × 1 | 720 × 731 × 1 | — | — |