International Archives of the Photogrammetry, Remote | 
P(F)= Y [P( F/L,)* P(L, )] 
where P(F) denotes the percentage of false defined polygons in 
total area, P(L;) the percentage of the landuse type within the all 
polygons and P(F/L;) the percentage of false defined polygons 
within the i" landuse type. By using the formula above, a total 
accuracy result of. 594.31 was determined. 
4. CONCLUSION 
As in all topics in geoinformatics, accuracy assessment Is an 
integrating part of a GIS. Without accuracy, GIS does only 
serve for limited aims. In this study, an accuracy assessment for 
both geometry and consistence was done. For this purpose, an 
external knowledge for geometric accuracy, an internal 
knowledge for thematic accuracy was used. The results on both 
geometric and thematic accuracies shows that the results of the 
landuse detection study can serve some disciplines such as 
urban and regional planning, transportation planning and 
management as well as facility planning and management. By 
interpreting the accuracy results, it must be taken into account 
that the scale of the landuse detection project was 1:25000. This 
will be a factor to accept the applicability of results for further 
studies. Moreover, this study shows the applicability of 
accuracy investigation for GIS's, which have big amount of 
record in database. 
Manual digitizing method let the work be harder. Furthermore, 
in some cases this causes the accuracy to be heterogeneous in 
whole project. But thinking on the richness and complexity of 
the legend for landuse classes, any of conventional automatic 
detection processes was not applicable. Moreover, unplanned 
and very complex situation in some parts of big cities like 
Istanbul, let the automatic method for line and texture detection 
not useful for landuse studies. This is a disadvantage for time, 
accuracy and cost. It is widely known that automatic methods 
have very big advantages, if suitable for the aim and if 
applicable. 
Sensing and Spatial Information Sciences, Vol XXXV, Part B4. Istanbul 2004 
REFERENCES 
Beard, K., Mackaness, W., 1993. Visual access to data quality 
in geographic information systems, Cartographica, 30-2,3, 37. 
45. 
Buttenfield, BP, 1993. Representing Data Quality, 
Cartographica, 30-2,3, 1-7. 
Celikoyan, T.M., 2004. Mointoring and Analysis of Landuse 
Changes in Historical Periods for the City of Istanbul by means 
of Aerial Photography and Satellite Imagery, , PhD Thesis, 1TU 
Institute for Science and Technology, Istanbul. 
Glossary of Cartographic Terms, 
www.lib.utexas.edu/Libs/PCL/Map_collection/glossary.html, 
accessed on 15.06.2003. 
ISO/TC 211 N 1034, 2001. Text for DIS 19113, Geographic 
information — Quality principles, as sent to ISO Central 
Secretariat for issuing as Draft International Standard, 
International Organization for Standardization, Oslo, Norway. 
Morrison, J., 2002. Spatial data quality, in Spatial databases : 
with applications to GIS pp. 500-516, Rigaux, P. Scholl, M. 
and Voisard, A., Morgan Kaufmann Publishers, San Francisco. 
Radwan, M.M., Onchaga, R., Morales, J.. 2001. A structureal 
approach to the amangement and optimisation of 
geoinformation processes, European Organisation for 
Experimental Photogrammetric Research, Official Publication 
No:41, Frankfurt am Main, Germany. 
Tastan, H., 1999. Data quality in geographic information 
systems, PhD Thesis, ITU Institute for Science and 
Technology, Istanbul. (in Turkish). 
1030 
KE 
ABS 
Sevi 
orde 
vari: 
Leic 
con 
agal 
betv 
appl 
diffi 
The 
HDS 
obse 
The 
puls 
scan 
dete 
retui 
The 
of th 
case 
that 
type 
atten 
poor 
com 
HL" 
althc 
have 
dista 
[2]. | 
on tk 
assui 
Wave 
good 
For 
accu 
accui 
mode 
mode 
Tuck 
pape 
matt 
retur