The accuracy of different methods for water body extraction in urban areas
| Serial number . | Method . | Kappa (%) . | Overall accuracy (%) . | Misclassification error (%) . | Omission error (%) . |
|---|---|---|---|---|---|
| 1 | Single-band threshold method based on B4 band | 94.06 | 99.89 | 6.51 | 5.23 |
| 2 | Single-band threshold method based on B6 band | 93.21 | 99.87 | 8.99 | 4.32 |
| 3 | Water index method based on NDWI | 65.66 | 99.14 | 48.29 | 8.54 |
| 4 | Unsupervised classification method based on ISODATA algorithm | 84.24 | 99.66 | 26.25 | 1.36 |
| 5 | Unsupervised classification method based on K-means algorithm | 76.79 | 99.46 | 37.1 | 0.56 |
| 6 | Supervised classification based on SVM | 91.63 | 99.85 | 1.31 | 14.35 |
| 7 | Decision tree model method based on multi-band combined operation | 94.41 | 99.90 | 4.42 | 6.6 |
| Serial number . | Method . | Kappa (%) . | Overall accuracy (%) . | Misclassification error (%) . | Omission error (%) . |
|---|---|---|---|---|---|
| 1 | Single-band threshold method based on B4 band | 94.06 | 99.89 | 6.51 | 5.23 |
| 2 | Single-band threshold method based on B6 band | 93.21 | 99.87 | 8.99 | 4.32 |
| 3 | Water index method based on NDWI | 65.66 | 99.14 | 48.29 | 8.54 |
| 4 | Unsupervised classification method based on ISODATA algorithm | 84.24 | 99.66 | 26.25 | 1.36 |
| 5 | Unsupervised classification method based on K-means algorithm | 76.79 | 99.46 | 37.1 | 0.56 |
| 6 | Supervised classification based on SVM | 91.63 | 99.85 | 1.31 | 14.35 |
| 7 | Decision tree model method based on multi-band combined operation | 94.41 | 99.90 | 4.42 | 6.6 |