8 8 ug duBamd ag 8073038008 
«em BLA 
en ludis i 5 EL ; 
ES 
N i AL 
Ld DIS 
4100 S e Te es 
IN Tu 
NS A. 
T CE 
  
  
  
  
  
  
  
  
  
  
  
  
Figure 1 : Tiling the HEIGHT and CLUTTER data 
       
  
   
  
CSI ID d 
PU CT 
e 
  
     
AC NON TER STE ES 
SUN Cette > CAM FAS 
  
    
  
  
18E - 24W 
  
  
Table 1 : WGS84 parameters 
2.3 Data Used 
For the generation of the products the following data were 
used: 
1. Panchromatic SPOT stereo pairs (P) from SPOT 
IMAGE catalogue. 
2. Multispectral SPOT images (XS) selected from 
SPOT IMAGE catalogue. 
3. LANDSAT TM images selected from the 
EURIMAGE catalogue (for some Aegean Islands ) 
4. Where available 1:50,000 scale topographic maps 
of the Hellenic Geographic Military Service 
(HGMS). 
5. GPS ground control points for areas where 
1:50,000 topographic maps were not available. 
3. DESCRIPTION OF THE PRODUCTS 
3.1 Height Database 
There were two options for the production of the Height 
Database. 
« SPOT STEREOPAIRS 
e 1:50,000 SCALE TOPOGRAPHIC MAPS 
916 
RT BAe FI CI EM 
AUCH CEA ET Ce 5 ry : 
D d RAO z DE] 
rh, os I: Aro 8 R328 1 Fo REN P d 
The SPOT. STEREOPAIRS digita! elevation model was 
computed by automatic correlation of SPOT stereo pairs 
in five (5) steps: 
1. Correlation of each pair Two (2) images of a 
SPOT stereo pair representing a stereo model 
were matched by automatic correlation. The result 
was analysed, validated and archived. 
2. Registration of all pairs covering the zone of 
interest: All pairs were registered from identified 
points on the images whose UTM co-ordinates 
were extracted from 1:50,000 scale topographic 
maps ( if maps were available) or from GPS 
measurement. In the North area of Greece where 
1:50,000 topographic maps were not available, 
GPS technique were used for determination of the 
ground control points. About 60 ground control 
points were measured (1cm accuracy). 
3. Individual digital elevation model production, one 
for each pair: Each disparity map was projected 
with the geometric modelling functions obtained 
from the previous step. A digital elevation model 
was obtained. 
4. Individual digital elevation models were inspected, 
adjusted and validated. 
5. Individual digital elevation models were combined 
into mosaics to make one digital elevation model 
at 20 meter resolution with UTM zone 34 co- 
ordinates. 
In some cases where SPOT_STEREOPAIRS were not 
available and in cases where there were some “holes” on 
the DTM (cloud or snow coverage etc.) 1:50,000 SCALED 
TOPOGRAPHIC MAPS were used. The DTM was. 
obtained by extracting the contour lines from the film of 
the 1:50,000 topographic maps. 
The absolute accuracy of digitised HEIGHT DATA is 
depended on the contour interval between topographic 
map contour lines. It was estimated as half of the contour 
interval (10 meters height accuracy for 1:50,000 scale 
topographic maps). 
The production line process for this product was : 
e Scanning the film contour lines from 1:50,000 
topographic maps at 250 dots per inch (dpi) 
resolution (5 meters on the ground). 
e Contour line extraction by automatic vectorization 
and manual height attribution. 
e Black and white scanning of 1:50,000 scale 
topographic maps at 250 dpi resolution (5 meters 
on the ground). 
e Rectification of maps in UTM 
cartographic co-ordinates. 
(zone 34) 
e. Spot height digitization. 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B4. Vienna 1996