nbul 2004
1. During
ed digital
nsiderable
asht (a),
ponding
yreliminary
nd spectral
ed on the
rator.
be defined
iodified by
ue. For the
samples as
ig data set
the more
be defined
ess. To test
ess for the
mples were
nces for the
ed on these
nembership
'apabilities,
) method a
lected. The
ts with the
mplexity as
xf our ACD
ction of the
lity of our
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B2. Istanbul 2004
Figure 3. 1:1000 planimetric map of the city of Rasht (a),
corresponding IKONOS Pan-sharpen Patch (b),
Corresponding Aerial Patch (c). Extracted 3D objects in test
area (d), Classified Objects (e) and Final result of proposed
automatic change detection methodology in the test area (f).
4. CONCLUSION
The obtained results by applving our proposed strategy on
different kinds of objects from natural to man-made GIS
objects established the high capability of our proposed ACD
strategy. The main feature of this strategy is not so much its
individual modules that perform different tasks, but the
methodology itself that governs the entire system. Our
methodology is based on these premises: (1) Simultaneous
fusion of all available information for the object extraction
and recognition. In our case these were limited to the three
STS components. However, it can be extended to include
other possible descriptive attributes if they are available. (2)
Because of the fuzzy behavior of the objects. a rigorous and
crisp modeling approach for extraction and recognition
problems should be avoided. (3) Taking into account the
numerous varieties of the objects types and appearances,
training potentials are a real necessity for an ACD method.
(4) Within the general scope of the proposed methodology,
individual modules such as matching operation, surface
modeling, region growing, structural and textural analysis,
491
etc. can be improved parallel with the related algorithmic
developments.
We believe our proposed ACD strategy has demonstrated a
promising and comprehensive solution to a complicated
problem, however, we are still far from reaching to a perfect
solution for a fully automatic ACD system. Bearing in mind
the general concepts presented above we may outline the
future research works based on the following proposals:
e Implementation of a hybrid neuro-fuzzy approach
by which recognition parameters as well as fuzzy
rules are trained and modified.
e Algorithmic improvements should be investigated
for individual modules govern the extraction,
recognition and reconstruction phases.
5. REFERENCES
Armenakis, C., Cyr, L, Papanikolaou, E., 2002. Change
Detection Methods for Revision of Topographic Databases.
Symposium | on Geospatial theory, Processing and
Applications, Ottawa.
Dowman, L, 1998. Automated procedures for integration of
satellite images and map data for change detection.
IAPRS,V ol. 32, Part 4, pp. 162-169.
Gonzalez, R.C., Woods, R., 1993. Digital image processing,
Addison-Wesley Publishing, Reading. Massachusetts.
Kim, J. R., Muller, J-. P., 2002. 3D reconstruction from very
high resolution satellite stereo and its application to object
identification. Symposium on Geospatial theory, Processing
and Applications, Ottawa.
Peled. A. 1993. Change Detection: First step toward
automatic updating. ACMS-ASPRS. Vol. 30, Part 4, pp. 281-
286.
Schiewe, J., 2002. Segmentation of high-resolution remotely
sensed data - concepts, applicaüons and problems.
Symposium on Geospatial theory, Processing and
Applications, Ottawa.
Shi, Z., Shibasaki, R., 2000. GIS database revision — the
problems and solutions. The International Archives of the
Photogrammetry, Remote Sensing and Spatial Information
Sciences, 32, Part B2, pp. 494-501.
6. ACKNOWLEDGEMENTS
The authors would like to acknowledge Iranian Remote
Sensing Center (IRSC) for providing IKONOS image,
furthermore much valuable help has been given by Mr. M.
Talebzadeh the Deputy of IRSC through the provision of the
IKONOS imageries, and National Cartographie center (NCC)
for providing 1:1000 scale 3D digital maps of the test area.
