Research Article
Remote-Sensing Image Tile-Level Annotation Based on Discriminative Features and Expressive Visual Word Descriptors
Table 2
The PA value comparison between using different methods.
| Methods | Water | Vegetation | Residential area | Farmland |
| Color features | 0.967 0 | 0.909 1 | 0.288 1 | 0.916 5 | Texture features | 0.938 2 | 0.678 8 | 0.377 7 | 0.551 2 | Spatial features | 0.883 6 | 0.948 3 | 0.550 6 | 0.726 8 | NOF + K-means | 0.929 0 | 0.937 6 | 0.612 0 | 0.787 1 | WOF + K-means | 0.934 9 | 0.949 4 | 0.616 8 | 0.815 7 | Our proposed method | 0.939 4 | 0.947 0 | 0.664 1 | 0.911 2 |
|
|