——— —Q
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV
. Part B5. Istanbul 2004
the control point is simultaneously recorded by photography
(Fig. 9 and Fig. 10).
images, find same objects in the field, and measure them as
geodetic references.
In case terrestrial photographs are used for documentation of
large areas, the use of targets becomes impractical for various
reasons. Since the scale of each photograph will be different
with respect to any location on site, the size of each target
should be adjusted respectively. Because of varying topography,
it is hard to control the visibility between the targets and the
images beforehand, whereas natural control points can be
selected from images after the photography
Figure 5. Survey images are based on concentric image
sequences. The images are transformed to the normal case of
stereoscopy and rectified on planes. The structures are measured
from stereo images. The angle of view of this panoramic
The evident payoff caused by the use of natural control points is
that they are less accurate than the geodetic and targeted ones.
However, improving the photogrammetric block geometry can
compensate this deficiency. This can be done by various means, composition corresponds to 92° x 67° horizontally and
like by increasing the number of control points in image, by vertically. |
increasing the number of images in block, or, by using |
concentric image sequences, just to name few.
2. METHODS
2.1 Panoramic photography
The survey images are recorded as wide area stereo
photography (Fig. 5). The photography is usually performed as
panoramic stereo image sequences. The practical location of the
camera is the opposite slope of the wadi. Each sequence
consists of two to three single images. For panoramic
photography, we apply concentric image sequences (Haggrén et
al., 2004).
A sequence of three images with an overlap of approximately
30 % corresponds to an approximate horizontal angle of view of ) ; eu |
20° S -2 A i x 7 C s: . 2
tm E P me been 15 5 meters and ; lar a RENS on Figure 6. Archaeologists use photogrammetric documents as
i EC E EE M P se as rave their notebook during field invention. The structures to be |
cen typically x meters. In general, we aim at base-to- measured are identified and sketched on prints. |
distance ratios between 1:10 or 1:15.
: |
cm image resolution. On upper parts of the slopes, where the A
photographing distances may exceed 300 meters, additional :
closer photography is necessary.
The camera, which we have used for the documentation, is res |
Olympus Camedia C-1400L. The size of a single record is 1280 f. d |
x 1024 pixels, and we use primarily the wide end of the zoom Fre wall |
optics, which corresponds to a focal length of 1400 pixels. The mm
photographing distances of 100-200 meters correspond to 7-14 Bade |
|
|
|
|
For the sake of occlusions caused by terrain topography, or by
bushes and trees, and in order facilitate detailed interpretation
of the barrage and terrace structures, we have additionally
recorded these by close-up stereo photography. These images
are recorded within a distance of some ten to twenty meters with
a base of 1-3 meters. The close-up stereo photography consists
mainly of single images, not of panoramic sequences.
Figure 7. Since no targeted control points are used, recognition
of natural features becomes essential part of the fieldwork of
photogrammetric documentation. Proper candidates are marked
here with triangles. The remains of structures are identified
The survey group has identified barrage systems and terrace during archaeological survey and marked on prints (here shown
in white). |
walls on marked them on paper prints (Fig. 6 and Fig. 7).
|
2.2 Control survey
As geodetic control points we have used natural features like |
corners or edges of large stones. The points are selected on
prints. They are then identified in the field and the coordinates |
are measured with tacheometer (Fig. 8). The exact location of |
