'anbul 2004
pdating process
or aerial photos
:d to detect and
badly damaged
/ (the center of
| 49? East and
lope lower than
clative to other
city have been
of 1:25000 base
quite outdated,
er provinces.
"M coordinate
control points
'€ used as input
contours, water
extracted from
211 defined and
vas collected
ps and on the
ations of these
xtracted from
DEM) of each
pixel in RMSE
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol XXXV. Part Bl. Istanbul 2004
and finally HRG image was orthorectified. In figure 5, topographic
vectors overlaid on the orthorectified image can be seen.
Figure 5. Overlay of vectors on the orthorectified image.
6. MAP REVISION STAGES
After preparation stage, such as orthorectifying satellite image based
on old topographic maps and mathematical model, old digital map
superimposed on the orthorectified image. Then 2.5m resolution
panchromatic image and ground data collected from the field
submitted to one image-interpreter. She worked with drawing
software and carried visual interpretation. on that image in a
monoscopic way. Changes were defined and grouped in four codes in
change detection phase (field work). Features that were deleted
relative to the old topographic maps, added features, geometric
changes and descriptive changes were reserved in the first, second,
third and fourth code respectively.
The existing following features exist in the study area were
considered and evaluated:
= Communication Lines such as roads, railways, etc and
associated features e.g. railway stations, bridges, etc;
= Administrative features including buildings, walls, stadium,
airport, Tunnel, Pool, Storage and fence;
* Power lines and pipelines such as electrical, gas and water
supply features and associated features e.g. water well, oil
or gas well;
=» Hydrological features, including rivers, streams, canals and
floodways;
= Land cover and vegetation, including cultivated areas, wet
lands, orchards and etc.
The interpreted features were grouped by layer.
7. CONCLUSION
Major problem which effects use of satellite images for topographic
maps revision lies in the low resolution of the satellite images and
therefore in the information content of the resulting revision maps.
The acceptable images should provide information content required
for revision of 1:25000 scale topographic maps.
By using orthorectified images, updating old topographic maps is
possible. Image processing software with the ability of superimposing
vector on raster has been used to carry out this operation.
Results of feature extraction showed that there were no difficulty to
detect and identify area features such as towns, smaller villages and
isolated buildings. All linear features such as roads, tracks, railways,
etc were extracted easily, except in some places where the contrast
was relatively low. Finally, the point features such as water / gas
wells, single trees were impossible to be detected and identified on
the image. For each layer, quality criteria according to the
interpretation results are defined as follow (total = 100%):
+
Un
i fe | i i :
0 me e p Hr ESL E mem T
& = 8 5 8 g z oO ê e UR 2 € o 82 E 2
2 = = = 5 = = = S : P © 3 o > = ©
= w S m 3 e sS = zu a = 5 Ii
= 0 t£ tü e I o = 9 3 o =
d = © ee o
3 g i
Figure 6. Results of detection and identification of features on SPO1
5 HRG image for 1/25000 scale map revision.
Changes can be detected based on comparison between aerial photo
acquired on 1991 with SPOT HRG image acquired on 2003. Rivei
>
ge.
Figure 7. SPOT 5 HRG image acquired on 2003.
For each feature type, detected changes with respect to the old
topographic map sheets are defined as follow (total 100%):
1004 35
80
6041
i
ge
0 te 2 2 = = © = > z v 9
rT 9 7 9 9 9 3 9 s E Ÿ E $9 S * S
8 E à 8S c 0 à S > 8 É 2.2 72 &
S2 2 = 8 3 «8 85 9 $S $ à 2 $.2
= Li =: 2 n c U 32 5 = 9» >
= o = m = n 2 5 = >
i z 3 = 9 o O e
3 & iL
Figure 8. Detected changes from 1991 until 2003.
