Conclusions
The method applied allowed to obtain an
accuracy of 1-3 cm in object coordinates at
the photographic scale from 1:40 to 1:100.The
image distance and the scales used are the
most significant in sea water where in
average conditions of visibility the very high
adsorption and scattering, that reduce the
light transmission, do not permit an
appropriate illumination and resolution for an
iobject distance more than 3-4 m.
The accuracy obtained is good , in fact the
values are near those declared in terrestrial
photogrammetry standard application with
non-conventional systems.
The data aquisition is relatively simple. It is
sufficient to use non-metric cameras and a
reference frame at a low weight and of
reduced sizes. Using only few points ( 8 in the
experiments) self-calibrating and bundle
adjustment reduction program it is possible
to obtain the tested accuracy.
There was not a significant difference in
accuracy using different focal lenght. The
cameras did not take photos exactly at 2 m.
distance and the photograms with different
cameras had not been taken in the same
environmental conditions, but with different
refractive index.
So we should conclude that we can
reconstruct in 3-D coordinates an object at
image distance approximately fixed, using a
number of variable photograms and changing
the focal lengths and the sea water
conditions. The variation in accuracy is not
very significant ( 1-3 cm.), relatively to
standard underwater applications. We have
only to collocate a reference frame in the
field of the photograms without using control
points on the object or on the structure.
The method presented could be employed to
reconstruct the form and the dimensions of
particular undersea object or structure. In
particular it is interesting when the subject
has a long extension in one dimension, for
example to survey an archaeological site or to
map the sea bottom.
The fig. 6 shows some indicative profiles in
a frontal and in a lateral view of the cave and
of the stalactite of photograms of fig. 5.
These profiles have been obtained in Autocad
utilizing the X,Y,Z coordinates of the
photogrammetric reduced points.
This paper showed the results of first
experiments of the method, some others
experiments with different systems and
methodology are going to execute, which
results will be published successively.
240
Acknoledgements to Geotop s.r.l. ( Ancona ,
Italy ) for Rollei MR2 photogrammetric
reduction system loan.
References
Fryer, G., Fraser, C.S., 1986. On the calibration
of underwater cameras. Photogrammetric |
Record. 12(67):73-85.
Hohle, J., 1971. Reconstruction of the
underwater object. Photogrammetric
Engineering. 37 (9):948-954.
Kotowski, R.,1988. Phototriangulation in
Multi-media photogrammetry. International
Archives of Photogrammetry . 27 (V ): 324-
334.
Leatherdale, J.D., Turner, D.J.,1991.
Operational experience in underwater
photogrammetry. ISPRS . Journal of
photogrammetry and remote sensing. 46:104-
112.
McNeil, G.T.,1969. Underwater photography.
Photogrammetric Engineering. 35(11) :1135-
1152.
Karara,H.M.,(Editor),1989. Non-topographic
photogrammetry. American Society for
Photogrammetry and Remote Sensing,Falls
Church, Virginia:445.
Pollio, J.,1971.Underwater
photography and sonar.
Engineering. 37(9) : 955-968.
Torlegard, A.K.I.,Lundalv, T.L.,1974.
Underwater analytical system.
Photogrammetric engineering.40 ( ):287-293.
mapping with
Photogrammetric