PolSAR Image Classification on Flevoland (test)
99.65Water AccuracyDDF2Pol
Evaluation results
| Method | Links | ||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| DDF2PolNetwork Domain=Dual-Domain CNN2026.04 | 99.65 | 99.52 | 98.49 | 95.54 | 97.96 | 93.55 | 96.7 | 99.29 | 99.26 | 98.45 | 98.2 | 98.55 | 99.6 | 98.3 | 100 | 98.16 | 98.2 | 97.99 | |
| 3D-CNNNetwork Domain=Real-Valued CNN2026.04 | 99.05 | 98.45 | 96.97 | 84.84 | 91.23 | 82.55 | 71.2 | 98.15 | 98.77 | 25.82 | 85.44 | 98.31 | 97.53 | 98.78 | 100 | 89.7 | 88.47 | 88.73 | |
| CV-2D-3DNetwork Domain=Complex-Valued CNN2026.04 | 98.41 | 93.42 | 97.49 | 87.44 | 81.95 | 89.22 | 94.16 | 95.56 | 98.45 | 98.13 | 96.42 | 91.81 | 95.1 | 97.76 | 100 | 93.6 | 94.35 | 93.01 | |
| HybridCVNetNetwork Domain=Complex-Valued CNN2026.04 | 97.55 | 99.6 | 97.81 | 97.84 | 94.69 | 93.84 | 90.14 | 98.81 | 99.6 | 97.55 | 96.45 | 99.21 | 99.59 | 97.22 | 100 | 96.7676 | 97.3313 | 96.4711 | |
| WaveletCNNNetwork Domain=Real-Valued CNN2026.04 | 97.25 | 97.63 | 98.63 | 83.92 | 95.65 | 91.54 | 77.84 | 96.4 | 99.39 | 98.89 | 96.47 | 84.29 | 99.12 | 99.39 | 99.48 | 93.78 | 94.39 | 93.21 | |
| PolSARFormerNetwork Domain=Real-Valued CNN2026.04 | 97.04 | 97.22 | 97.22 | 90.29 | 94.48 | 91.76 | 84.7 | 98.42 | 96.81 | 97.97 | 90.67 | 95.95 | 97.44 | 98.29 | 100 | 94.41 | 95.22 | 93.9 | |
| 3D-CV-CNNNetwork Domain=Complex-Valued CNN2026.04 | 95.68 | 99.32 | 91.38 | 72.98 | 94.34 | 90.71 | 87.28 | 90.4 | 98.47 | 95.79 | 88.57 | 88.5 | 95.12 | 98.05 | 98.79 | 92.17 | 92.36 | 91.45 |