3.2. Influence of ground resolution 
A few test series with different ground resolutions of the 
optoelectronic system were carried out. There are two 
ways to realize it. The first one is to change the flight 
height, the second one is a change of the field of view with 
a constant number of pixels per line. Both ways yield the 
same results. 
It could be established that the influence of the camera's 
ground resolution is not as drastic as the influence of the 
area's elevation dynamic. 
Consequently it would be better to reduce the data rate of 
a camera by using macro pixels and not by using a greater 
compression factor. The blocking effects caused by certain 
compression algorithms make it harder for the match algo- 
rithm to find conjugated points, so the number of matched 
points decreases. The elevation error caused by the lower 
number of matched points could be much greater than the 
error caused by a smaller ground resolution of a camera. 
In general the results of the previous chapter could be 
established. 
3.3. Other aspects 
Another point to plead for a small stereo angle is that the 
stereo processing just can be started, when all lines got 
information about the same area. Consequently the 
greater the stereo angle the bigger the memory storing the 
image and position data. 
In contrast to the simulation we have to take into consider- 
ation a bi-directional reflectance in reality. The smaller the 
stereo angle the smaller the differences of images caused 
by this effect. 
Maybe a more intelligent and more expensive match algo- 
rithm is able to match a sufficient number of points even if 
the stereo angle is large. But actually there is no need for a 
great stereo angle, because the accuracy to determine a 
single point's elevation will not be improved. 
4. CONCLUSION 
The main results are: An optimal stereo angle exists and 
should be between 10 and 40 degrees. The permissible 
range depends on the elevation dynamics of the observed 
surface and on the camera's ground resolution. A stereo 
angle between 15 and 20 degrees leads to a sufficient 
number of matching points and to a good accuracy of the 
elevation values under all conditions. 
These results underline the importance of a suitable simu- 
lation technique for the design and optimization of camera 
parameters and observation conditions. 
5. REFERENCES 
Bórner, A., 1995. Optimierung optoelektronischer Sys- 
teme; Diplomarbeit, Technische Universität Ilmenau 
Heipke, C., 1994. Digitale photogrammetrische Arbeitssta- 
tionen - Algorithmen, Design und Bewertung; Habilitations- 
schrift, Technische Universität München 
Reulke, N., 1995. Simulation und Optimierung optoelektro- 
nischer Systeme am Beispiel der Bestimmung von 
30 
Wolkengeschwindigkeit und -höhe; Dissertation, Tech- 
nische Universität Berlin 
Reulke, R., Reulke, N., Jahn, H., 1994. Numerical simula- 
tion system for the generation of image data from space- 
borne imaging sensors for planetary exploration; SPIE Vol. 
2318, pp. 144-154 
U.S. Geological Survey, 1992. Mission to Mars: Digital 
Topographic Map, Volume 7: Global Topography 
WAOSS, 1991a. Wide Angle Optoelectronic Stereo Scan- 
ner (WAOSS), Mars-94 Mission, Phase B Study, WAOSS 
Science Objectives; Berlin 
WAOSS, 1991b. Wide Angle Optoelectronic Stereo Scan- 
ner (WAOSS), Mars-94 Mission, Phase B Study, WAOSS 
Technical Part; Berlin 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B1. Vienna 1996 
Abstı 
GPS | 
Statio 
frame 
Progr 
numb 
The | 
opera 
such 
kinen 
execu 
navig 
Those 
pecul 
proce 
satell 
Here 
exper 
1. Int 
The » 
NAV 
low s 
in sig 
The : 
condi 
data 
throu 
proce