International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B7. Istanbul 2004
carries a state-of-the-art sensor and is the most advanced
commercial imaging system in orbit.
One of the pictures of this satellite belonging to our project
area is shown in Fig. 2.
F 7E e
QuickBird
dr. BÉ Tobi om o A. i
Figure 2. image in Riyadh
2.3 Ground Control Points (GCPs)
In the orthorectification of HRS images, GCPs play very
significant role. For IKONOS and QuickBird images, GCPs
were established to dm-level accuracy at a spacing of about 5
km. In case of LANDSAT-7 images, GCPs were established
to m-level accuracy at a spacing of about 50 km.
In order to establish GCP archive to be used in future, for
each GCP, a description card was prepared and in many
cases photos were also taken at these points from a short and
medium range. The coordinates of GCPs were determined in
the national datum (Shedayed et al., 1992) as well as the
WGS84 datum.
2.4 Digital Elevation Models (DEMs)
It is to be noted that the IKONOS Geo product and
QuickBird Basic and Standard Ortho Ready Imagery Data
requires to be processed with the refined Image Support
Data (ISD) set, including the orbital parameters of the sensor,
GCPs and a Digital Elevation Model (DEM), to obtain the
[mage accuracy at the level of 1.0 — 1.5 pixels.
The accuracy and resolution of the DEM required for the
orthorectification of the Satellite Images is depending on the
[mage georeference position accuracy specified by the
Customer. Since it is aimed to have 1.0 — 1.5 pixel accuracy
in our projects, a DEM with an accuracy of at least 5 meters
is required.
2.5 Orthorectification
The system corrected satellite images are subject to errors
and distortions due to height differences and lack of precise
ground control points. These errors may be as big as 50
meters in case of the system corrected HRS images
depending on the magnitude of vertical angle and
topographic relief. As part of this project, the
orthorectification of IKONOS and QuickBird images to 1.0 —
1.5 pixels was performed very successfully. The
orthorectification process was carried out using Russian
OrthoSpace, which uses a dynamic model without requiring
sensor camera parameters and / or PCI with sensor
parameters and rational polynomial coefficients. The
mosaicking was carried out using IRAS-C software.
3. SAUDI GEODATABASE USING HRS
IMAGES AND PLANS
3.1 Image Geodatabase
Using HRS images, a very comprehensive geodatabase was
established. It consists of about
100,000 km? of IKONOS and QuickBird images (Fig. 3)
2,000,000 km” of LANDSAT-7 images (Fig. 4)
structural plans of the entire Saudi Arabia
- local plans and subdivisions of Qaseem, Riyadh and
Aseer regions (Fig. 5)
= in.
Mes
eds =, 2 den
J =. - *
Pa —— MÀ ;
"A 3 . ^
ks .
A B k Fo
1 » " e i
% P * ^i
T a = ? - "
, deem "i n
5 bs ; 4*3
i Tn
* " 8 o. i
i N Ha
4 V. i
1, LEG amem
s. ^ u
^ M Pag mm. oe i r
pe Rr, ——
Figure 3. IKONOS & QuickBird coverage (100,000 km2)
Figure 4. LANDSAT-7 and structural plans coverage
(2,000,000 km2)
Inter:
The 4
space
À gec
landu:
image
The |
using
postin
geom
is pro
solvin
A sea
orthor
Arabi:
image
data si
The
check
checki
pickec
were |
agains
values
The o
other :
REG
Najran
Gizan
Aseer
Baha
Hail
Arar
Riyadl
Jeddah
Damm
First, ]