anbul 2004 
IG 
have been 
ere found. 
non metric 
metric and 
tereoscopic 
à reduction | 
analytical 
[he overall 
than 3 cm. 
concern to 
otted using 
1318/10 
n a mobile 
ited with a 
e from the 
  
   
th the help 
jalls were 
ric (reseau 
cameras: 
ns (with a 
x 4/45 and 
1 Sigma 20 
gent cases. 
» reach a 
ration) and 
ect/camera 
with these 
th the help 
bh tripods) 
    
Figure 7 
International Archives of the Photogrammetry, Remote 
Sensing and Spatial Information Sciences, Vol XXXV, Part B5. Istanbul 2004 
  
Figure 5. Hasselblad installed on stere 
Analogue UMK and Hasselblad photographs were digitized 
with an Intergraph TD Photoscan at 1814 dpi, and Pentax 
images with a desktop scanner Nikon Super Cool SCAN 2000 
at 3000 dpi. 
^ 
+ 
  
  
  
33 Data reduction 
The data reduction methods employed in this project were 
stereoscopic 
conventional 
analytical 
B 
A b 
plotting, 
photogrammetry and rectification. 
3.3.1 Conventional analytical plotting. Combination of a 
metric camera (UMK 1318/10) analytical plotter and normal 
case has been used in the main facade and main external walls 
(in the east side). In these cases both elevation drawing lines 
and rectified images mosaics were done (Figures 6 and 7). 
1 
oscopic bar 
  
  
ces ~ 
softcopy 
Figure 6. Elevation view of the main facade and mosaic of 
rectified images (Canon D30 Digital Camera) 
illustrates 
the architectural lin 
es, 
cracks 
and 
indications of main disorder in the walls in the east side. These 
walls. show large deformation, with a dangerous trend to 
collapse. In fact, the walls have ties and they are shored up. 
  
   
Figure 7. Elevation view of the walls in the east side. 
3.3.2 Digital Photogrammetric Workstation (DPW). The 
study of the vault that covers the chapel (Figures 2 and 8) was 
made by zenithal shots organised in three parallel strips (5 
photos per strip). Camera positions were situated in the ground 
by means of a grid and helped by tripods. 
  
Figure 8. The Ortophotograph, contour map and main profiles 
showing the deformation in the chapel vault. 
The Canon D30 camera with the 20 mm lens was used. Focus 
was set to infinity. Although the camera had been calibrated in 
laboratory (by selfcalibration with minimal inner constraints 
and digital target location; Atkinson, 1996; Cardenal et al., in 
this volume), an additional selfcalibration was performed with 
the block of 15 photographs in order to refine the inner 
parameters in real conditions. Image point measurement was 
manual. Control and check points were used and calculations 
were carried out with a routine programmed under IDL? 
(Interactive Data Language from Research System Inc.). Inner 
parameters were: focal length, principal point offset and K, and 
K, (first and second radial distortion coefficients). Other 
distortion coefficients (higher order terms or decentering