Istanbul 2004 
  
3M2 
the 90% of 
1. 81% of 
nd that the 
us that the 
is assumed 
evaluation 
rea. In this 
created by 
S images, 
9 43°, The 
is obvious. 
re which is 
ns that it is 
areas and 
'he critical 
about 30°. 
possible to 
it is very 
  
Figure 4. Example 1- HRS back image. 
As a general conclusion, the accuracy of the DEM which is 
produced by HRS using only the metadata information for the 
orientation process, is better than 10m in mountainous areas, 
except in areas where the along track slope is larger that 30°. 
6. FURTHER WORK 
In the field of sensor modelling the main object is to improve 
the generic along track stereo satellite model with respect to 
the Inertial Coordinate system in a way that: 
e The information which is extracted from the metadata 
will be used not only to solve the model but also to refine 
and improve the solution. 
e The self calibration process will be examined deeply and 
also, the role of the acceleration and the angular velocity. 
In the field of DEM generation the next steps are 
e Extract a DEM using the UCL software and the UCL 
model for the DEM orientation. 
e If available, use nadir HRG images of Pitkin area with 
almost the same acquisition date as HRS images, in order 
to test how we can improve the DEM using the 
combination of HRS and HRG images in areas with 
along track slope larger than 30°. 
* Improve matching accuracy by investigating new 
algorithms. 
7. SUMMARY and CONCLUSIONS 
This paper has described the testing of the UCL sensor model 
and the generation of a DEM using ERDAS Leica 
Photogrammetry Suite software (beta version). The results 
that are introduced within the paper guides us to the following 
conclusions: 
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV , Part Bl. Istanbul 2004 
e The SPOT HRS data can be oriented using ground 
control points or metadata and that a solution can be 
found which is within the expected error bounds. 
Especially, the accuracy of the heights compared with the 
GCPs is very close to 5m. 
* use of self calibration gives a slightly improved solution. 
e The accuracy when only using orbital data is good, and 
that the solution with ground control points is probably 
constrained by the accuracy of the control. 
e The almost simultaneous acquisition time of the HRS 
images is the key to achieve high correlation during the 
image matching process better than 90%. Comparable 
figures for SPO I’ HRV are around 82%. 
e The accuracy of the sensor model and the high 
correlation of the image matching are the two principal 
factors of getting the expected DEM accuracy. 
e The accuracy of the DEM which produced by HRS using 
only the metadata information for the orientation process, 
is better than 10m in mountainous areas, except of the 
areas where the along track slope is larger that 30°. 
Finally, in general, SPOTS-HRS shows that the use of the 
along track stereo sensors is a very promising for DEM 
generation. 
8. REFERENCES 
Airault S, Gigord P, Cunin L, Breton E, Bouillon A, Gachet, 
Bernard M, 2003. Reference3D location performance review 
and prospects. ISPRS workshop on High Resolution mapping 
from space, 2003, Hanover. 
Baudoin A., Schroeder M., Valorge C., Bernard M., Rudowski 
V. “HRS-SAP initiative: A scientific assessment of the High 
Resolution Stereoscopic instrument on board of SPOT 5 by 
ISPRS investigators". ISPRS workshop on High Resolution 
mapping from space, 2003, Hanover. 
Leica 2003. *Leica Photogrammetry Suite:User Guide". 
SPOT Image, 2002. “Spot Satellite Geometry Handbook", S- 
NT-73-12-SI Edition 1 Revision 0,15-1-2002 
SPOT Image, 2003. "DIMAP Dictionary version 1.1- SPOT 
Scene profile" - Edition 01 September 2003 
Valorge C., 2003 “3D Restitution and rendering through high 
resolution imagery: State of the art and new challenges”. 
ISPRS workshop on High Resolution mapping from space, 
2003, Hanover.