International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B1. Istanbul 2004 
  
The stereo pair CE histogram of the stereo pairs has a median 
value of 6 meters and a mean value of 7.1 meters. The 90-th 
percentile of the stereo pairs CE sample is 11 meters. The 
same skew pattern observed in the monodrop comparison is 
present in the stereo comparison. There is a noticeable 
improvement in geolocation accuracy over the unadjusted 
monodrop images. The act of relative triangulation helps 
reduce some of the effects of random error. 
  
atitude (meters) 
e 
| 
1 
= 
  
  
  
235720715510 S 0 5 10 15 20 25 
longitude (meters) 
Figure 12: Stereo Pairs Random Bias 
The 90-th percentile of the stereo pairs random biases 
sample ranked according to radial distance from the origin is 
8 meters and is shown as the blue dashed line in Figure 12. 
  
0.1 
0.05 
  
  
  
0 10 20 30 
LE 
(meters) 
Figure 13: Stereo Pairs Linear Error Histogram 
The stereo pair LE histogram of the stereo pairs has a median 
value of 8 meters and a mean value of 9.1 meters. The 90-th 
percentile of the stereo pairs LE sample is 16 meters. The 
same skew pattern observed in the monodrop and stereo 
pairs CE comparisons is present in the stereo pairs LE 
comparison. 
6. DISCUSSION 
Obtaining the maximum system performance for a complex 
system like OV-3 takes place over a period of time. As the 
system is exercised, improvements are made in system 
models, tuning and operation of the system. The largest 
improvements in system performance are expected to occur 
early in the program and later improvements are expected to 
be incrementally smaller. Towards the end of its life, the 
components may degrade and system performance can suffer. 
It may be difficult to predict the exact shape of the 
performance curve in advance. Figure 14 shows a nominal 
expectation of the system performance over time. 
The OV-3 system performance has exceeded the level of 
operational maturity and additional improvements in 
performance are expected. 
  
operational 
  
  
1 1 | 1 
  
time 
Figure 14: Expectation of System Performance Over Time 
7. CONCLUSIONS 
In the above sections the calibration software and 
calibration range used to perform the on-orbit geometric 
calibration of OV-3 have been described. The on-orbit 
geometric calibration of OV-3 has proceeded through a 
series of steps concluding with the geometric camera 
calibration. System level tests using comparisons to ground 
check points have validated the operational geolocation 
accuracy performance and the stability of the calibration 
parameters of OV-3. 
REFERENCES 
Kohm, K. and Tira, N., 2004, On-Orbit Image Quality and 
Radiometric Characterization of the OrbView-3 High 
Resolution Imaging Satellite, Accepted for publication in: 
Proceedings of the ASPRS 2004 Annual Conference, May 23- 
28, 2004. 
Kuang, D., 2004. Evaluation of the OV3 Orbit Accuracy. JPL. 
Mulawa, D., 2000. Preparations for the On-Orbit Geometric 
Calibration of the OrbView 3 and 4 Satellites, The 
International Archives of the Photogrammetry, Remote 
Sensing and Spatial Information Sciences, Amsterdam, Vol. 
XXXIIL Part BI, Book 1, pp. 209-213. 
MOD 
KEY WO] 
ABSTRAC 
The Modu 
used in ren 
function. A 
MTF isa r 
the along s 
accurate m 
statement C 
The classic 
obtain MT] 
nominal 0] 
measureme 
The data s 
Results are 
1.1 Motiv 
On-orbit qi 
desirable 
commissio! 
requiremen 
the end use 
quality of 1 
equalize th 
in a combir 
This work 
OrbView-3 
throughout 
measureme 
use of de 
metric of th 
While dedi 
results, acc 
nominal : 
opportuniti 
estimate o 
imagery wi 
array. This 
the linear a 
required w 
complemen 
1.2 Objec 
The objecti 
measureme 
on the orde