opaiou
t aerial
aim of
for the
ping of
ations,
rder to
cessed
> were
un the
bes the
image
:nce of
r from
North-
72002.
pplied:
ial and
High
juently
ion is
n of a
rea. A
in the
digital
large-
om an
o face
metric
Ice an
of this
etailed
often
ng out
by the
:k and
erived
ith the
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B5. Istanbul 2004
*
5
x
8
3
&8
mu
Figure 2. The old map of the city drawn by K.R. Lepsius (1843)
with superimposition of the GCPs derived from GPS survey.
2. GPS SURVEYS
The applications of GPS in Archaeology are numerous and refer
to the determination of the absolute position of reference points,
the detail survey of single structures, the kinematic surveying of
the ground surface of entire sites. The reliability of this global
system and its possibility to achieve high accuracy, together
with its productivity, make it the main technique able to frame
together the whole metrical information collected at the site in a
common reference system (Bitelli and Vittuari, 2000).
Regarding Soknopaiou Nesos site, long time GPS observations
were firstly conducted on two reference points established in the
area, connecting them to the IGS permanent stations of Matera
and Noto (Italy), Nicosia (Cyprus) Ankara (Turkey) and
Malindi (Kenia); GPS allowed to link the coordinates of local
reference points to the International Terrestrial Reference Frame
ITRF2000, and moreover to insert previous surveys, referred to
local reference frames, into ITRF. The accuracy of the absolute
positioning came out at the sub-decimetre level, certainly
redundant compared to the common requirements of
archaeological surveying.
At the same time, Stop-and-Go method was adopted to obtain
the coordinates of 59 — signalized and natural — ground control
points (figure 2), to be used in photogrammetric exterior
orientation (Bitelli et al., 2003a).
In order to realize a general Digital Terrain Model (DTM) of the
entire area, kinematic surveying was finally conducted,
following paths chosen to describe the morphology of the site.
About 20000 points were measured with a good accuracy (in the
order of some centimetres). Figure 3 shows the planimetric
scheme of GPS measured profiles.
From the interpolation of the points acquired, a complete DTM
of the site was produced (figure 4).
The orthometric heights were obtained applying the geoidal
undulation values obtained from the EGM96 geoidal global
model.
3. PHOTOGRAMMETRIC DATA PROCESSING
As mentioned, images were acquired by a low-height aerial
photo-acquisition system specifically developed for this kind of
applications (Bitelli et al., 2001): the system is based on balloon
or kites and a radio-controlled device. During the survey
conducted in February 2002, a 35 mm non metrical Canon
EOS500N film based camera was adopted instead of a
photogrammetric camera. The camera was calibrated by a field
calibration procedure (Bitelli et al., 2003a).
In relation to the area extension, the acquisition system
characteristics and the wide angle 24 mm lens adopted, a flight
height of about 100 meters was planned, involving an image
scale of about 1:4000. Considering the archaeological
requirements and the organizational and logistic constraints, this
value of photo-scale can constitute an acceptable solution to
have an high level of detail in the products provided by digital
photogrammetric processing (DTM, orthophotos, base layer for
GIS establishment).
32634004
Figure 4. 3D view of the DTM obtained by GPS survey.
Strong wind during the survey and the consequent irregularity
of balloon flight path, made the numerical remote control
inadequate to impose the defined camera orientation. In total
about 140 photos were acquired with a resulting coverage
geometry characterised by irregular overlap and camera tilt
besides large image scale differences.
Selected images were digitised by a Wehrli RasterMaster RM2
photogrammetric scanner with maximum geometric and
