anbul 2004
| camera; a
S purpose,
) x 2.043
in order to
and on the
was least-
| the outer
st circular
parameters
adius. The
"Rec, and
arameters.
ed for the
inder after
ch control
idjustment
'stablished
is, it was
"formation
neans that
:e and the
the lineal
'8 m. The
best fitted
ited.
rammetric
aling with
orientation
he object
recting the
lowing the
d images.
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B5. Istanbul 2004
NELTITIMEMDIITANENOIL MENTI IRRBNCIIUMUBACIISONREANTLCONÜVRE IIT PUR ZEMENT LUNG IE
Caption
EEE TTT ETT ETT TTT
fon asa as: 252-143 an 0 98 1:022 13 3 35 2 43 3 A3 6.
Figure 4. Error map showing discrepancies between both
surfaces (real vs. modelled).
Furthermore, this program performs the development of both
inner and outer surfaces, depending on the direction i.e.
leftwards or rightwards. Figure 5 shows the result of two
outer developments.
Figure 5. Set of two unwrapped images
33 Results
The final product was a raster digital product that shows the
entire tower's exterior surface (Figure 6). The final image is
obviously scaled, allowing it to be used in multiple tasks:
inventories, visualisation and analyses of monument
deteriorations, measurements (punctual, lineal or areal),
updating of architectural information systems, plottings,
thematic mapping. It is even worthy for foreseeing future
conservation and restoration actions, and many multimedia
purposes of our Cultural Heritage can be derived.
Besides, it was also estimated the final accuracy of the digital
product in order to be sure whether or not the initial
geometric requirements were fulfilled. For this purpose,
many measures were computed, considering control and test
points belonging to the tower. Mean discrepancies,
minimum, maximum and relative errors, as well as mean
Square errors were computed.
Following the direction of development, i.e. the x direction,
the relative error in distance was 0.1696, meanwhile in the
transverse direction, i.e. the y direction, the relative error was
approximately 0.19%. Obviously, errors in both directions
mean different, the former refers to the best cylindrical
adjustment error; the latter depends overall on the surveying
measurements.
Figure 6. Mosaic of the whole exterior of the small tower
4. CONCLUSIONS
This paper has dealt the mapping of mathematical surfaces
corresponding to architectural or archaeological objects,
using photogrammetric unwrapping techniques.
After analysing the results, it can be estimated that the
methodology of raster projection and development of
mathematical surfaces is right. Furthermore, it should be
considered whenever regular objects are concerned.
Moreover, the availability of metric scaled images of the
whole object or monument, in our case the small tower, eases
the documentation, analyses, catalogue production and
conservation of our Cultural Heritage. Furthermore, the
visual impact and multimedia possibilities of mapping 2D
and 3D objects should be emphasised.
Besides, this methodology only requires the knowledge of a
small sample of control points, and conventional digital
images. Thus, the combination of this monoscopic measuring
technique and the employment of camera calibration
procedures based on the image space make expensive the use
of DesenRec 1.0 to similar targets. Following this way, the
measurement of planar and curved objects does not depend
on the digital camera used, and it solves many problems
coming out in mapping exterior and interior architectural or
archaeological sites or monumets. Furthermore, this
technique complements and widens the way of traditional
stereo or monoscopic mapping.
5. REFERENCES
Hemmleb, M., Wiedemann, A., Digital Rectification and
Generation of Orthoimages in Architectural
Photogrammetry. International Archives of Photogrammetry
and Remote Sensing, Goteborg, Sweden, Vol. XXXII, Part
SCIB, pp. 261-267, 1997.
Karras, G. E., Patias, P., Petsa, E., Ketipis, K., Raster
Projection and Development of Curved Surfaces.
International Archives of Photogrammetry and Remote
