If a mesh is projected onto an object by a projector having the
same optical characteristics as camera employed, the
photogram of the object with the projected mesh (raster-
photogram)and the pseudo-photogram constitute the equivalent
of a stereoscopic pair. Actually each projected point has an
image on the raster-photogram (fig. 3) and an image on the
pseudo-photogram (fig. 2). It is therefore possible to obtain the
coordinates of the points projected on the object. This
restitution is analytical and requires a specifically designed
program which reconstructs the model from the plate
coordinates of each point, measured on the raster-photogram in
monocular observation, and from the coordinates of the same
point on the pseudo-photogram.
Figure 3 Portion of a Raster-photogram
The accuracy of the monocular reading is directly related to the
sharpness of the lines of the mesh. In fact this "sharpness"
should be correlated to the device employed for the monoscopic
measurements. The pseudo-photogram employed, which was
already existing as it had been previously used for another
survey, has the following characteristics:
Two types of vertical lines,
(pseudologarithmic) spacing:
a. Principal lines, 0.08 mm thick, equally spaced 10 mm
apart and identified with marks every 20 x 20 mm
b. Secondary lines, 0.04 mm thick, of variable spacing.
Starting from the centre, the sequence of intervals is as
follows:
S(1)=1+1x(+0.05;+0.04;+0.03;+0.02;+0,01;-0,01;-
0,02;-0,03;-0,04;-0,05)
for the first centimetre,
S(n)=1+1x n x(0,05-1)
for the ith space of the nth centimetre.
Two types of horizontal lines:
having variable
24
a. Principal lines, 0.08 mm thick, equally spaced 10 mm
apart and identified with marks every 20 x 20 mm
b. Secondary lines, 0.04 mm thick, equally spaced 1 mm
apart.
The plan for the raster survey is shown in fig. 4; the altar to
camera distance is approximately 2.30 m for 3 of the sides of
the altar and 2.60 m for the West side. The distances were
conditioned by the presence of steps around the altar which
made it impossible to place an object the size of the projector
any closer. The projector/camera base about was 400 mm. We
tried to make normal exposures with a horizontal base parallel
to the altar and the axes of the camera and of the projector
normal to the base.
The bas-relief is so detailed that a greater photogram scale,
that is a shorter camera to object distance, would have been
appropriate.
The projected points of the mesh are separated, on the altar, by
a horizontal and vertical distance of approximately 2 cm. As
can be observed in fig. 3, since the surface relief of the altar is
fairly small, the points are rarely not visible; it is also
uncommon for the background to hinder the points' visibility
because of it's darkness.
More than 3000 points were projected on the smallest side of
the altar, the North one.
3.2 Second Phase
As mentioned previously, the second phase, which has only
just begun, aimed at plotting points inside the mesh. For this
purpose two stereo raster-photograms were necessary. Actually
the 2 photographs were taken with the two cameras positioned
symmetrically to the middle of the altar while the position of
the metric projector was central to the altar. In fact, adhesive
targets were placed on the edges of each side of the altar: these
targets were connected to an external topographic network and
made it possible to check the scale of the model and to relate
the surveys of the four sides of the altar to the same reference
system. The two stereoscopic raster-photograms made it
possible to perform the restitution of the points which were
inside the mesh Figure 4 shows the arrangement used.
4. TAKING PHOTOGRAM
Problems arose due to the reflections produced by the golden
or silver surface of the leafs, which were only partially solved.
Tests regarding lighting position were made in order to reduce
reflections. Several exposures were made with films of varying
sensitivity. PEY64 film was employed, generally with a
diaphragm setting of 11 and exposure times varying from 3 to 6
minutes. All exposures had to be made at night because the
Abbey was being used during the day. The film was assured to
be unshrinkable but in fact was found to exhibit some
deformation in time.
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B5. Vienna 1996
podus, uo
10322f od
| BI3UIBO
BIIUIEBI
The software
creation of th
collinearity ec
relative to the
In particular tl
each projected
nodal point of
by the same p
point of the cai