International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B4. Istanbul 2004
occurrence of high-order respectively significant nodes, in-
depth study of the type/shape of the nodes. To get on without
predefinition of thresholds or the preliminary fixing of minimal
and maximal values it is our aim to continue the search for
broader regularities, for example in the combination between
above mentioned criteria.
Regarding this we could imagine many hypotheses and to find
the following or similar structures with data mining:
- capitals are always located at large rivers?
- in general all big cities are located at large rivers?
- jn the city centre are larger buildings than in
outskirts?
- in tourist areas are more bicycle tracks than in non-
tourist areas?
- industrial areas are situated mostly along big traffic
routes?
- winding roads are always in regions with heavy
differences in elevation?
- villages are embedded mostly in agricultural crop
land, very rare they are located in forest?
- .. 90 per cent of all junctions of traffic lines are situated
in settlement areas?
We will concentrate on both ways, supervised and unsupervised
methods. Both can support knowledge discovery and during the
implementation of algorithms, both data mining models will
influence each other.
5. CONCLUSIONS
The paper presented attempts in the range of spatial data mining
in the context of realising a spatially aware search engine. To
solve spatially related queries, the computer has to be aware of
semantic aspects. Ontologies are used to represent them.
However the information therefore can not completely be
acquired manually. Automatic detection and learning processes
of the computer are essential to enrich such data collection.
Classical metadata are a first approach to reveal the content of a
data set. Our intention is to extract metadata automatically from
geographical data sets. An automatic enrichment with specific
metadata, e.g. the keywords, was presented.
Further steps are necessary to make semantic of geographical
data visible, so that the computer receives background
knowledge and can perform logical reasoning procedures.
Therefore we use and implement data mining methods. In this
article concepts and first attempts were introduced and
explained, which have emerged as main focus during our
investigations. First algorithms were developed and realised.
6. REFERENCES
Anders, K.-H., 2003. Parameterfreies hierarchisches Graph-
Clustering Verfahren zur Interpretation raumbezogener Daten.
Dissertation Universität Stuttgart.
Berners-Lee, Tim, James Hendler & Ora Lassila, 2001. The
Semantic Web. Scientific American.
Brenner, C., 2000. Towards fully automatic generation of city
models. Proceedings of the XIX ^ ISPRS Congress, volume
XXXIII-B3 of /nternational Archives of Photogrammetry and
Remote Sensing, pages 85-92, Amsterdam.
Elias, B., 2002. Automatic Derivation of Location Maps.
IAPRS Vol. 34, Part 4 "Geospatial Theory, Processing and
Applications", Ottawa, Canada.
Gerke, M., Butenuth, M., Heipke, C., 2003. Automated Update
of Road Databases Using Aerial Imagery and Road
Construction Data, IntArchPhRS, Munich, Vol. XXXIV Part
3/W8, pp 99-104.
Heinzle, F., Kopczynski, M., Sester, M., 2002. Spatial Data
Interpretation for the Intelligent Access to Spatial Information
in the Internet. Proceedings of 21st International Cartographic
. Conference, Durban, South Africa.
ISO/TC-211, 2003. Text for FDIS 19115 Geogaphic
information - Metadata. Final Draft Version. International
Organization for Standardization.
Jones, C.B. et al, 2002. Spatial Information Retrieval and
Geographical Ontologies: An Overview of the SPIRIT project".
Proc. of the 25th Annual International ACM SIGIR Conference
on Research and Development in Information Retrieval,
Tampere, Finland pp.387 — 388.
Koperski, K. & Han, J., 1995. Discovery of Spatial Association
Rules in Geographic Information Databases, im M. J
Egenhofer & J. R. Herring, eds, ‘Advances in Spatial Databases
95°, Vol. 951 of Lecture Notes in Computer Science, Springer
Verlag, Heidelberg, pp. 47-66.
Lillesand, T.M., Kiefer, R.W., 1994. Remote Sensing and
Image Interpretation. John Wiley & Sons, Inc. New York -
Chichester - Brisbane - Toronto - Singapore.
Michalski, R.S., Bratko, [., Kubat, M., 1998. Machine
Learning & Data Mining: Methods and Applications. John
Wiley, England.
Schenk, T., 1999. Digital Photogrammetry, Volume 1. Terra
Science, Laurelville. 428 pages.
Sester, M., 2000. Knowledge acquisition for the automatic
interpretation of spatial data. [International Journal of
Geographic Information Science. Vol. 14, No. 1, pp. 1-24.
Straub, B.-M., 2003. Automatic Extraction of Trees from
Aerial Images and Surface Models. ISPRS Conference on
Photogrammetric Image Analysis, September 17-19, Munic,
Germany, IntArchPhRS, Vol. XXXIV, Part 3/W8, pp. 157-164.
Thomson, R., Richardson, D., 1999, The 'good continuation'
principle of perceptual organization applied to the
generalization of road networks. Proceedings ICA '99, Ottawa,
Canada, Session 47B.
Witten, LH., Frank, E., 2000. Data Mining, Practical Machine
Learning Tools and Techniques with Java Implementations.
Morgan Kaufmann Publishers.
7. ACKNOWLEDGEMENT
This work is supported by the EU in the IST-programme
Number 2001-35047. We thank the National Mapping Agency
of the state Lower Saxony in Germany (LGN) for providing the
ATKIS data and the National Mapping Agency of France (IGN)
for providing several topographic data sets.
340
KEY
ABST
A met
may h
If ther
model
a com
The p
Satelli
applic
envirc
manag
terms
Storag
netwo
satelli
Mittal
On th
applic
essent
chip i
correc
establ
GCP:
be coi
time ¢
of the
If sat
geom:
neces:
match
image
strean
Becat
param
and (
methc
In thi
match
There
match
circun
inforn
comp!
match
