- Clusters post-processing, A win python program (https://winpython.github.io/) consisting of 3 modules clu_r.py and the 2 library MODULEs clu_data_headers, & clu_myf.py, that perform:
- Visualization of feature space with 2-d plots per cluster
- Visualization of feature space with 3-d plots per cluster
- Statistics of clusters (min, max calculation)
- Linear regression of 2-d feature spaces (eg. 3 combinations for ALOS, SRTM, ASTER GDEMs) per cluster
- Video: https://vimeo.com/258236125
- Data: Cluster image and the feature space images derived from related projects (SVR & SVR.DEM). The cluster image is named MASK.TIF (0= no data, 1, 2 ... for cluster classes). The feature space images (eg. pc2, pc3 reconstructed elevations) used in clustering are named 01.tif, 02.tif, etc. The vector data model in SVR.CLUSTERS differs from SVR & SVR.DEM projects, since the first column indicate a) 0 for no-data, and b) 1, 2 ... for cluster classes, while the next columns correspond to image data files (01, 02, 03 ..), for example the residual elevations of ALOS, SRTM, ASTER GDEMs.
**Figure a2.** _2-d feature space visualization per cluster (7 clusters)_
Figure a3. Clusters centroids (7 clusters)
Figure b1. Visualization of the reconstructed LST.
Figure c1. Selected 3d scattergrams per cluster.
Table c3. Min, max statistics per cluster.
| NBG | ALOS | (m) | SRTM | (m) | ASTER | (m) |
|---|---|---|---|---|---|---|
| Clusters | Min | Max | Min | Max | Min | Max |
| 3 | -183.3 | -8.2 | -166.8 | -7.4 | 18.5 | 350.8 |
| 5 | -8.2 | -0.9 | -7.4 | -0.7 | 4.6 | 18.5 |
| 6 | -0.9 | 1.9 | -0.7 | 1.8 | -0.7 | 4.6 |
| 1 | 1.9 | 3.7 | 1.8 | 3.5 | -4.2 | -0.7 |
| 2 | 3.7 | 5.4 | 3.5 | 5 | -7.3 | -4.2 |
| 4 | 5.4 | 7.5 | 5 | 6.9 | -11.2 | -7.3 |
| 7 | 7.5 | 13 | 6.9 | 11.9 | -21.8 | -11.2 |
| 3 | 13 | 335.9 | 11.9 | 305.9 | -634.7 | -21.8 |