International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B5. Istanbul 2004
2.2.3 Digital processing: First, camera calibration
characteristics are entered into the PhotoModeler software
program. Then, the images are downloaded into a computer
system and ingested into the software as well. Tie points, or
ground control points, had to be identified in the pictures that
would allow the pictures to be tied together. The operator
identifies these points and records them. It is necessary that
each of these points appear in at least three of the photos. Once
a sufficient number of points are identified and labelled, the
automated photogrammetric process can be run. At this point,
the software can calculate the measurements and distances
between points and also determine camera locations for each of
the photos, building a virtual three-dimensional model.
3. APPLICATION
The methods outlined above were employed in two
experiments. The first application was in the documentation of
the interior of a modern building. This was done to develop
methods for application in an archaeological setting. After this
experimental site was recorded and documented, the refined
methods were applied to the Salinas monument.
3.4 Hibben Center Experiment
The first part of the experiment involved using a digital camera
to acquire pictures in a modern building to establish the
methods necessary for photo-documentation of a cultural site.
The building used was the Hibben Center for Archaeological
Research Building on the University of New Mexico campus.
Its high-ceilinged atrium-like construction provided a suitable
simulation for documenting large and tall structures. The
interior space of the building is approximately 16 by 10 meter
with a 13 by 7 meter atrium open to the third floor. The
perimeter ceiling height of the first floor is 2.7 meters high,
while the ceiling of the atrium is 11 meters high. This space
provided an excellent example of tall architecture, and provided
a simulated experience for documenting large structures in a
confined space.
The camera used for recording this site was the Nikon 990 with
an FC-ES8 fish-eye lens. The experiment involved taking a
minimum number of pictures that would still allow for the
photogrammetric reconstruction of a three-dimensional model.
A total of ten pictures were collected from ground level. The
camera was set to macro focus, and the lens setting to extreme-
wide angle. The full wide-angle capability, the fish-eye setting,
was not used, as this had proved too difficult to calibrate in the
photogrammetric software. The extreme wide-angle lens was
capable of capturing wall's full height and length from all
angles, meaning that 11m high walls could be captured in their
entirety from a standard tripod height. Using a constant camera
height of approximately 1.5 meters, a straight-on image was
captured for each of the four walls, pictures were taken from
each of the corners, from two ends of the building and two
points near the center of the building facing each end. (Figure
3)
326
Hibben Center for Archaeological Research Bldg. , First floor -
showing perimeter ceiling overhang and atrium opening to third floor
yu X | pz
A
À > A T A^ 4
Y First Floor Perimeter Celling Height - 2.7m [
4 rt sunl
pu , Open to Atrium ceiling, height - 11,5 m
i
i + i - "1
4. ;
1
i
!
1
| Eee eee nee aes
B. Y ;
4 A AN 4. 4.
deem E
A scamera position and angle
10 meters |
L ER 4 3 4 3
Figure 3. Hibben Center Building layout showing the camera
placement and viewing angles for each picture.
Each of the pictures captured the wall of interest in its entirety
and additional area beyond. These last two camera positions
allowed the capture of the entirety of the building, as the fish-
eye lens covers 183 degrees; each picture captured slightly
more than one half of the building interior. In all cases, the 15
meter high walls were captured in their entire height, and each
straight on picture covered the 18 meter length of the walls,
eliminating the need for excessive photos. A minimum number
of shots were necessary, however, to capture every detail of the
building in at least three photographs in order to tie points
together to build the model.
The digital images were then downloaded to a computer and
ingested into the PhotoModeler photogrammetric modeling
software application. Building the model involved “stitching”
the pictures together by selecting common points, or ground
control points, between the photos. Each point used had to
appear within a minimum of three pictures to be useful for
constructing the model. Once ground points were selected, the
program was run to build a three-dimensional model. The
software utilizes camera calibration information along with
photogrammetric algorithms to compute distance and
measurements in the photos. With this, the application can
build a virtual three-dimensional model with approximate
dimensions (Figure 3). The software utilized outputs a model
that can be imported into drafting software applications for
documenting layout and dimensions.
3.2 Salinas Monument Experiment
The methods developed in the Hibben Center experiment were
then applied in an archaeological setting. The mission church
in the Quarai complex of the Salinas National Monument was
chosen for its size and structure. The walls of the mission
church still stand in some places at a height of 12 meters. It has
been a challenge for the National Park Service to record this
structure sufficiently, as access to these heights are difficult
without causing damage to the walls and structures.
The camera utilized in this application was a Nikon Coolpix
5700 with a FC-E9 fish-eye lens in the wide-angle setting. The
St
al
pl
WwW
SU
ta
CC
m
Sur
for
at :
det
caf
adı
of 1
The
Ph
Hit
con
tog
that
Sur!
