Therefore the following
measuring instruments were used: an
electronic tachymeter (ELTA 2)
combined with an electronic
interface (REC 500) and a small
plotter, and a (non-metric) survey
camera (ROLLEIFLEX SLX) with a 50
and 80 mm lens. The potential of
this type of photogrammetric camera
was often used for architectural
tasks, e.g. Stephani M., Eder K.
1987, Wester-Ebbinghaus W. 1983.
This equipment offered a high
measurement comfort and
considerably facilitated the
outdoor work.
The determination of control
points was carried out by polygons
for the outside walls and the
interior area of the the castle.
The differences in elevation
between these two parts (about 25
meters) and bad sights rendered
this work rather difficult. In some
parts also tachymetric measurements
were carried out for further data
processing in order to minimize the
expense of raw-data collection.
The photogrammetric work can be
divided into two parts: all walls
were covered by terrestrial camera
stations and in addition -
especially for the interior area -
aerial photographs were taken using
a helicopter of Aeroflot.
PHOTOGRAMMETRIC TASKS AND DATA
PROCESSING
The photogrammetric tasks were
given either by the object and the
desired result. Because a mapping
of. the borders of the stones
principally had to be guaranteed,
the image scale should not be less
then 1:300, and for a better
visibility color images were taken.
The accuracy of the graphical
representation should be in the
range. Of 0.05-0.1 m. This was
achieved for both mentioned types
of photo coverage. It must be added
that the calculations containing
aerial photographs had worse
accuracy because of movement blurs.
control points
points per
The number of
was. restricted. to 6
wall, because after sorting and
ordering all pictures at first for
almost each sight of wall the
number of control points was
increased by bundle block
adjustment Düppe R. D. 1984.
Afterwards stereosscopic models
were evaluated with an analytical
plotter (Wild AC1). The data
collection was carried out due to
242
required entities containing
different types of lines which were
supplied with different keys due to
their information (e.g. profile,
detail, contour). It was attached
importance to get all relevant
information (line) simultaneously
evaluating a model. After having
finished this the raw-data for the
following processing with AutoCAD
were given in a homogeneous
coordinate system.
DATA PROCESSING WITH AUTOCAD
A brief look at the literature
shows that AutoCAD is more and more
used for the graphical
representation of photo-
grammetrically derived data, e.g.
Belli A. 1990, Heine E. et al.
1990, Stevens D., Mc Kay W. M.
1990. The advantage of this
procedure is that data sets of
different origin can be combined.
The user has also a lot of
utilities for visualisation
available | and the exploitations
were done on a graphic monitor.
Before transfering and
integrating the data into AutoCAD a
preprocessing had to be carried out
with regard to the following
aspects:
connection of line segments
belonging together derived from
different models
combination of data sets which
belong to one plot
transformation of the lines into
the right projection plane
cleaning of the data
This steps were finished with
the conversion of the data into the
DXF-format which means a "data
exchange format" for in- and output
data for AutoCAD (s. AutoCAD User
Guide). The tachymetric data were
already given in the DXF-format,
but only stored as single points.
Therefore these points were
connected interactively using a
rough Sketch drawn at Amberd
castle.
For the postprocessing
different information was put into
varying layers to enable to fade
data in or out due to the
respective task. Thus the input
could be controlled by e.g.
superimposing a front view with the
vertical profiles of a wall. The
graphical representation could be
varied by signature, colour or area
filling if necessary. The result
should be an easily readable
representaion of the entire
information (in fig. 2 an overview
of all plans is depicted).
