4 Results
An unsupervised classification based on 10 endmembers was performed on the
CHRIS/PROBA dataset (Figure 3).
Figure 3. Unsupervised classification result performed on the CRIS/PROBA
image showing the potential of the hyperspectral image for coral reef
classification.
Class names could be assigned using the ground truth information collected
during the field campaign. This classification result should not be considered as a
highly accurate bottom-type classification. However, it shows that, based on the
spectral information contained in the CHRIS/PROBA dataset, different bottomtypes
could be automatically distinguished. Moreover, the patterns detected are
not random noise, but show clear structures consistent with the morphology of
the reefs.