The impacts of vegetation indices from UAV-based RGB imagery on land cover classification using ensemble learning

The production of land use and cover (LULC) maps using UAV images obtained by RGB cameras offering very high spatial resolution has recently increased. Vegetation indices (VIs) have been widely used as an important ancillary data to increase the limited spectral information image in pixel-based classification. main goal this study is analyze effect frequently RGB-based VIs including green leaf index (GLI), red- green-blue vegetation (RGBVI) triangular greenness (TGI) on classification images. For purpose, five different dataset combinations comprising bands were formed. In order evaluate their effects thematic map accuracy, four ensemble learning methods, namely RF, XGBoost, LightGBM CatBoost utilized process. Classification results showed that with increased overall accuracy (OA) values all cases. On other hand, highest OA calculated Dataset-5 (i.e. considered). Additionally, result Dataset-4 TGI) superior performance compared Dataset-2 GLI) Dataset-3 RGBVI). All all, TGI was found be useful for improving having GLI RGBVI. improvement reached 2% index. Furthermore, within algorithms, produced (92.24%) consist RBG considered.