ibul 2004 
ingle of 
se four- 
attitude 
tellite's 
images 
he flow 
3. SIMULATION EXPERIMENT 
Two strips was designed, each strip has two satellite images. 
The strips have longitudinal and lateral overlap. And the four 
stimulant images were created (see figure 4, the shadow is the 
overlap area of the adjacent images), the attitude and the 
adjacent images' pixel coordinate and geodetic coordinate were 
known. 
  
  
  
  
  
  
  
  
Figure 4. Stimulated area images 
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B-YF. Istanbul 2004 
Considering the computation precision of the attitude and the 
targets’ ground coordinate are mainly influenced by image 
points’ measured error, the experiment contains two instances. 
The first is to analyse the influence on the attitude computation 
precision that image point error caused with varying overlap 
measure, e.g. the overlap measure is 10%, 20%, 30% or 60%. 
The second is to analyse the influence on the targets’ poisoning 
precision that image point error caused when different overlap 
was given, e.g. the overlap measure is 10%, 20%, 30% or 60%. 
The first instance: the analysis of influence on the attitude 
computation precision that image point error caused with 
varying overlap measure. 
The experiment was divided into four groups, i.e. longitudinal 
overlap is 10% & lateral overlap is 10%, longitudinal overlap is 
20% & lateral overlap is 20%, longitudinal overlap is 30% & 
lateral overlap is 30%, and longitudinal overlap is 60% & 
lateral overlap is 20%. The measured error is variable from 0.00 
mm to 0.03 mm. The difference the terrain is 2000 m. the result 
of this experiment is as follow. 
  
  
  
  
  
  
  
  
  
Measured error Pitch error (second) Roll error (second) Yaw error (second) 
0.00 mm 0.0022-0.2467 0.0000-0.0016 1.2566-5.6549 
0.01 mm - : 1.7158--4.7654 0.6713-3.4740 12.4537-17.7531 
002mm 3.4269-9.5125 1.4262-7.0223 30.1944-36.3754 
7 70.03 nm : | 5.1427-12.2407 2.2660-10.64427 47.5632-54.6107 
  
Table 1. relationship between the measured error and attitude error 
(longitudinal overlap 10% & lateral overlap 10%) 
  
  
  
  
  
  
  
  
  
Measured error Pitch error (second) | Roll error (second) Yaw error (second) 
: 0.00 mm 0.0022-0.0247 0.0000-0.0016 1.2566-5.6549 
0.01 mm. 0.2633-1.0049 0.1352-1.1020 2.6934-6.8807 
; 0.02 mm _ 0.5206-1.9830 0.3072-2.1623 10.2831-12.6474 
"0.03 mm | 0.7100-2.9589 0.4547-3.1792 15.2730-20.2182 
  
Table 2 relationship between the measured error and attitude error 
( longitudinal overlap 20% & lateral overlap 20%) 
  
  
  
  
  
  
  
  
  
Measured error Pitch error (second) Roll error (second) Yaw error (second) 
0.00 mm 0.0022-0.0247 0.0000-0.0016 1.2566-5.6549 
001 mm 0.2916-0.7645 0.2429-1.7177 0.9321-7.2375 
DO mm. 0.5973-1.5023 0.4820-3.4197 0.8062-17.9389 
— -0.03 mm ; 0.9073-2.2380 0.7170-5.1059 0.8790-28.3340 
  
Table 3. Relationship between the measured error and attitude error 
(longitudinal overlap 30% & lateral overlap 30%) 
  
Roll error (second) 
Yaw error (second) 
  
0.0000-0.0011 
1.2566-5.6548 
  
0.0118-0.2255 
10.6213-15.0465 
  
0.0242-0.4479 
17.5512-30.9677 
  
  
  
Measured error - Pitch error (second) 
0.00 mm 0.0021-0.0246 
00 mm 0.2655-0.9097 
0.02 mm . 0.5148-1.7936 
= 0.03 mm . E 0.7611-2.6761 
  
0.0373-0.6660 
  
24.5598-46.5067 
  
IH 
Table 4. Relationship between the measured error and attitude error 
(longitudinal overlap 60% & lateral overlap 20%)