Research Article
Google Earth Engine for Advanced Land Cover Analysis from Landsat-8 Data with Spectral and Topographic Insights
Table 3
Quantitative results of the suggested SVM approach for land cover classification.
| Method | Classes | Precision | Recall | F1 score | OA | Kappa |
| SVM + SB | Water | 0.9839 | 0.9899 | 0.9869 | 0.8412 | 0.7983 | High vegetation | 0.9428 | 0.885 | 0.913 | Low vegetation | 0.6773 | 0.6208 | 0.6478 | Crops | 0.6733 | 0.7179 | 0.6949 | Built-up | 0.8225 | 0.8944 | 0.857 |
| SVM + SB + SI | Water | 0.9886 | 0.9889 | 0.9887 | 0.8607 | 0.8230 | High vegetation | 0.9217 | 0.9261 | 0.9239 | Low vegetation | 0.6992 | 0.7026 | 0.7009 | Crops | 0.7241 | 0.7374 | 0.7307 | Built-up | 0.8782 | 0.8592 | 0.8686 |
| SVM + SB + SI + TF | Water | 0.9875 | 0.9932 | 0.9903 | 0.8847 | 0.8534 | High vegetation | 0.9546 | 0.9125 | 0.9331 | Low vegetation | 0.7514 | 0.7536 | 0.7525 | Crops | 0.7866 | 0.7878 | 0.7872 | Built-up | 0.8673 | 0.9107 | 0.8885 |
|
|