Data aquisition 
Photos have been taken at depth of about 15 
m. and in different environments ( sea 
bottom and sea cave.fig. 5,7,3) where the 
illumination conditions are quite different and 
the refractive index variation is different 
for the variation of physical parameters 
(pressure,temperature and salinity ). 
A nadiral photogram and two lateral 
convergent photograms have been taken to 
permit a better convergence of the multi- 
image orientation algorythm. (fig.5,7,8 ). A 
favorable relationship could be obtained 
between stereobase and photographs distance 
{ 1:3- 14 ). 
The imaging distance to the test frame for 
the three cameras position was approximately 
2 m., resulting in an average photoscale of 
about 4:100 ( 15 mmj,.1:50 ( 28 mm.) and 
1:40( 35. mm.) 
Photographic equipment 
The camera are two Nikonos lll with 15 mm. 
Nikkor-UW f/ 2.8 lens and 28 mm. Nikkor -UW 
f/3.5 lens and a Nikonos IVA 35 mm. Nikkor 
f/2.8 lens (fig.3). The light sources was 
fornished by electronic flash Nikonos SB 102 ( 
150 watts ) and Sea &Sea 200 (150-200 
watts ) (fig. 3 ). 
Negative 35 mm color film ( Gold Kodak) 
with 100:21 ASA/DIN was used. 
Fig. 1 - Frontal view of the reference 
frame utilized. 
  
235 
Reference frame 
The frame is used for two purposes: to 
fornish the known points for the orientation 
elements of the photograms and to execute the 
cameras self-calibration 
The reference frame is made of PVC bars 
which form approximately the edges of a 
parallelepiped of about 0.9 x 0.2 x O.15 m. in 
size and of about 3 kg in weight ( fig. 1,2 ). 
The known points are signalized with white 
circular discs , each of 30 mm. diameter with 
within a black cross printed, that were 
affixed to the structure to serve as targets. 
The positions of the targets were determined 
with an high accuracy (< 0.1 mm) in a 
reference system fixed on the frame . 
The PVC termal dilatation coefficient is 
about 7 ppm for C? The variation of 
temperature from the surface to the working 
area at a depth of 15 m. was about 5 C? (13 
C?versus 18 C?) that produces an eventual 
length variation of about 0.35 mm. for the 1 
m. maximum length of the bar. This variation 
is less than the expected accuracy target 
obtainable through photogrammetric 
measurements. 
  
Fig. 2 - Lateral view of the reference 
frame utilized.