Table 1 Parameters used in the object-based image analysis process
| Â | Segmentation parameters | Classification | ||||
|---|---|---|---|---|---|---|
Segmentation level | Spectral band1 (weight) | Scale | Shape | Compact ness | Type of classification | Object Features2 |
Level 1 | MS (1) | 200 | 1 | 0 | Boolean membership functions | NDWI: Mean |
Level 2 | MS (1) | 50 | 0.8 | 0.3 | Boolean membership functions | NDVI: Mean |
| Â | Â | Â | Â | Â | Nearest | G: Mean, SD |
| Â | Â | Â | Â | Â | Neighbour | R: Mean, SD |
| Â | Â | Â | Â | Â | Classifier | NIR: Mean, SD |
| Â | PAN (0) | Â | Â | Â | Â | PAN: Haralick Dissimilarity, Haralick Entropy, Haralick Homogeneity |