International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B4. Istanbul 2004
4. CONCLUSIONS
The increased spatial resolution of Quickbird imagery is a
powerful information source for the identification of buildings,
but the complexity of an urban scene makes it difficult to get a
satisfactory results based only on spectral information. On the
other hand, laser scanner data are rich in spatial information,
but poor on spectral information, which makes it difficult to use
only altitude data to separate object with the same height. In the
present study, the integrated use of both data sources was
presented. The great amount of data and the different nature of
the features require a different treatment than the classical pixel
based analysis. Therefore, the object oriented approach was
used. In the analysis of the segments, spatial features, like form
and texture parameters were not sufficient to classify buildings.
The best results were obtained using laser data, because the
objects can be well delineated in a nDSM. The separation of
trees and buildings can be performed using the spectral
information of the satellite image. In this manner, both data
cooperate and the best aspect of each data source can be used.
The resulting thematic map does not allow to identify each
building, but only blocks. The results encorage to use high
resolution imagery together with laser scanner data to locate
new buildings in urban regions. It ca be used in a second step,
where the geometry of the buildings can be reconstructed,
analyzing only a small part of the image arround the selected
regions.
5. REFERENCES
Antunes, A.F.B. 2003. Classificaçäo de ambiente ciliar
baseada em orientaçäo a objeto de imagens de alta
resolucdo espacial. Ph.D. Thesis. Universidade Federal
do Parana, Curso de Pósgraduacáo em Ciéncias
Geodésicas.
Baatz, M. and Schüpe, A. 2000. Multiresolution
Segmentation — an optimization approach for high quality
multi-scale image segmentation. In: Strobl, J., Blaschke,
T., Griesebner, G. (eds.): Angewandte Geographische
Informationsverarbeitung | XII, Wichmann- Verlag,
Heidelberg, 12-23.
Centeno, J.A.S., Antunes, A.F.B., Trevizan, S.J., Correa,
F. 2003. Mapeamento de áreas permeáveis usando uma
metodologia orientada a regiôes, € imagens de alta
resoluçäo. Revista Brasileira de Cartografia. Brasil,
v.55, n.01, p.48 - 56.
Centeno, J.A.S.; MIQUELES, M.A. Deteccáo €
caracterizaçäo ed vegetaçäo usando dados do laser
scanner aerotransportado. In: /// Colóquio Brasileiro De
Ciéncias Geodésicas, 2003, Curitiba.
AL
Haala N.; Brenner C. 1999. Extraction of buildings and
trees in urban environments. /SPRS Journal of
Photogrammetry and Remote Sensing, vol. 54, iss. 2, pp.
130-137(8).
Ivits, E. ; Koch, B.; Blaschke, T. And Waser Lars. 2002.
Landscape connectivity studies on segmentation based
classification and manual interpretation of remote sensing
data. In: eCognition User Meeting, 22-24 October, 2002,
München.
Maas H.-G.; Vosselman G. 1999. Two algorithms for
extracting building models from raw laser altimetry data.
ISPRS Journal of Photogrammetry and Remote Sensing,
vol. 54, iss. 2, pp. 153-163. Elsevier Science.
Weidner, U. and Fórstner, W., 1995. Towards automatic
building extraction from high resolution digital elevation
models. /SPRS Journal of Photogrammetry and Remote
Sensing . vol. 50 iss. 4, pp. 38-49. Elsevier Science.
6. ACKNOWLEDGEMENTS
The authors would like to thank the support of the Federal
University of Paraná and CNPq for providing the scholarship
and finantial support for the research, and also LACTEC and
Tese for the cooperation.
592
KE)
ABS
This
high
II an
Seco!
touri
were
cell;
In re
and €
persp
the c
main
prodi
the gi
time
large
Partic
obtaii
25i
comp
in m
intern
Vassi
There
can p
simila
This i
the us
orthoy
and Q
orthoy
were |
and fc
The ı
Sicilia
only f
Was u
by usi
ÁS ca
locate