Research Article
Extraction and Classification of the Supervised Coastal Objects Based on HSRIs and a Novel Lightweight Fully Connected Spatial Dropout Network
Table 1
Results of the overall effectiveness evaluation.
| | Methods | OA (%) | Kappa | mIoU (%) | Parameters (MB) | Loss coefficient |
| | U-Net | 96.34 | 0.73 | 76.96 | 252 | 0.0275 | | U-Net (CBS) | 98.12 | 0.87 | 87.34 | 252 | 0.0051 | | DeepLabV3+ | 98.33 | 0.88 | 86.77 | 65 | 0.0043 | | DeepLabV3+ (CBS) | 98.50 | 0.90 | 88.47 | 65 | 0.0028 | | LFCSDN (ours, BN) | 98.24 | 0.87 | 87.72 | 16 | 0.0040 | | LFCSDN (ours, CBS) | 98.22 | 0.88 | 86.72 | 16 | 0.0012 |
|
|
