ons observed from
th the total station.
5.
to digitize the stereo
If the stereo models
y, the outlines of the
llts. Also, drawings of
5 do not cause any
other details, such as
ry dependent on the
rule, the results lead
gist (“the expression
nt from the original”)
ou wantto map things
rather long revision
lise accepted by both
ed to 1 : 20.
Hand drawing.
nna 1996
Since vector drawings from photogrammetric imagery are
of great importance to archaeologists, we have thought of
several ways of overcoming the obvious problems:
- Contour lines can give additional geometric informa-
tion and improved perception (see figure 8, or a very
fine example in SONG and YAO, 1990), but additional
work is needed to draw the contour lines or to derive
them from a DOM.
- |f DOMs are available, a method might be developed
to extract vectors from the slope information. If suc-
cessful, this could lead to an automatic or semi-
automatic procedure forthe derivation of vector plans.
With the exception of outlines and sharp edges, we
could not yet formulate acceptable criteria for the
presence of a vector line, however. Consider a drapery
facing the observer. Which lines should be drawn?
The line on the ridge, or in the depression, or the line
where concave curvature changes to convex? Or all
of those?
- DOMSs can obviously be used for all kinds of per-
spectives and photorealistic representations. We
have stopped our efforts shortly before getting a result
showing a sculpture as it would appear on a photo-
graph, asking ourselves why we do not use a good
photograph in the first place.
- An artist, creating hand drawings, is more capable of
giving an appropriate impression of an object by using
hatching techniques and filling in supplements for
details that have vanished. If enough identical points
can be found, a transformation of the artists view to
the exact geometry of the photogrammetric plot can
show good results (figure 7). The transformation has
to be observed very carefully, however, in order to
avoid unnatural distortion.
In some cases, objects being partly destroyed still can be
reconstructed graphically without any assumptions.
Figure 9 shows a reconstructed column. The pedestal
which is still in situ was plotted from stereo models. The
shaft was added using hand measurements taken from
the near-by fallen over object. From the capital of the
column shown in figure 9, only one fragment showing one
of the eight edges was found. The reconstruction of the
missing parts was based on the photogrammetric plot of
the fragment, making use of the the symmetry of the
octogonal object. All procedures were carried out in the
CAD system.
3.7 Laser Scanners
The development of laser scanners for close range
applications (KLEINER and WEHR, 1993) may result in
an alternative to photogrammetry as soon as these
instruments are field reliable and have sufficient resolu-
tion. For further processing and visualization of the data
the same problems as with photogrammtric DOMs are
encountered.
4. CONCLUSIONS
A large variety of modern surveying techniques is avai-
lable for archaeological projects. The Tang emperors'
mausoleums survey may encourage more archaeologists
to consider the new methods for their tasks.
Figure 7: Head of stone sculpture, reduced to about 1 : 12.
Above: Photograph. Center: Overlay of photogrammetric
plan (thin lines) and supplemented hand drawing. Bottom:
Hand drawing transformed to photogrammetric plan.
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B5. Vienna 1996
