ass (black 
orrelation 
step 2). A 
oon show 
entially at 
z many so 
t images, 
gistration 
ific height 
ewing an- 
the initial 
from the 
Figure 2). 
sifications 
map” are 
the final 
choice of 
values as 
runs) one 
he NDVI- 
one clas- 
a first one 
cample for 
n improve 
 tempera- 
ssification 
  
  
  
Figure 6: Daedalus channel six image (0.695 — 0.750 um, recorded at 12:30 MET, flight altitude: 300 m) (left) and two vegetation 
maps calculated with the algorithm described above with different initial seeds. Areas which are classified as vegetation are shown 
in white. 
  
  
  
  
Figure 7: NDVI-image (recorded at 12:30 MET) (left) and vegetation maps resulting from applying certain thresholds on NDVI 
data; middle: pixels having NDVI-values greater 0.5 are assigned to the vegetation class, right: same method using a threshold of 
0.3 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXII, Part 7, Budapest, 1998 
233