Fig. 8 is what showed the proportional accuracy of X, 
Y and Z coordinates which is based on the change of 
convergent angle and number of control point(object 
distance, 1.0m). If the number of control point gets to 
be many, the accuracy gets to be improved, but the 
rate is reduced gradually. So, the number of control 
point which is suitable for the condition of object is 
required. When the convergent angle changes from 20° 
to 120° , the accuracy of X, Y coordinates is highest at 
40° ~50° , and than it is reduced gradually, but the 
accuracy of Z coordinates was improved in proportion to 
the size of convergent angle. 
Fig. 9 is what illustrated the r.m.s.e of X, Y and Z 
coordinates based on the change of number of 
photographs and convergent angle. According as the 
number of photographs increases into 2 sheets, 4 
sheets, 6 sheets and 8 sheets, the r.m.s.e of X, Y and 
Z coordinates is being reduced remarkably. In addition, if 
the convergent angle changes from 20° to 140° , the 
rm.s.e of X coordinates increases suddenly, and the 
rm.s.e of Z coordinates reduces suddenly, but the 
rm.s.e of Y coordinates is showing almost constant 
aspect. So, it can be seen that the r.m.se of Y 
coordinates is not influenced by the change of 
convergent angle. 
Fig. 10 is what showed the proportional accuracy of 
X, Y and Z coordinates based on the change of 
number of photograph, and convergent angle. If the 
convergent 
3.4 Convergent Angle and Number of Sheets of Photos 
  
  
  
  
  
  
  
  
  
X coord. 
  
  
    
    
      
Y coord. 
  
60 60 
40 — 
X or Y Coord. 
Simulated proportional accuracy (x 1,000) 
  
  
— 2 
20: — 20 
8 
/ 
; 4 
10 9 fe — 10 
o 
o 2 
Oo 
N A A S 
————— 
TYPICAL RANGE 
0- z 
  
  
| | | | | | | 
20° 40° 60° 80° 100° 120 140 
Convergent angle ( °) 
  
Fig.c2:10" -Proportional“~accuracy” of X, Y. and Z 
coordinates which is based on the change of number 
of sheets of photos and convergent angle. 
        
  
      
  
  
  
     
  
  
  
  
  
  
  
  
  
  
  
  
  
[ >, 
7] 
T TE 
PS 
450 tt | Aa CPC 450 ic I T y Ke 
[47 L7 LA P P] 400 ^ [^| E P 
400 E dela def IS «s 400 1 | ET] Ese 400 
ë € = Ede ET IA TS E S 300 € ef I P m 
5 3 300 LA LT 4471 P EN 3 30H | ri. [^ P 300 
- Ë 260 FLT sh £59 3 250 |] FL. Ease 260 
3 3 wot] | AT] POST 2o 3 200 ^| | : ET | EI TIS 200 
5 3 150 HL d SS 160 2 160 || er PN 150 
% | E Ps % I |] N > 
100 ^ | i Ds 100 100 |^] ; f M 100 
60 $e 50 [1 l| so 
9 S 
0 
Z coord. 
Fig. 9 R.m.s.e of X, Y and Z coordinates which is based on the change of number of sheets of photos and 
convergent angle. 
angle gets to be large, the accuracy of Z coordinates 
is improved suddenly. But, the accuracy of X, Y 
coordinates is being reduced gradually, while it shows 
the highest state at 40° . Thus, it can be seen that the 
optimum condition of convergent angle that the 
proportional accuracy of X, Y and Z coordinates is 
appearing most satisfactory is 80° ~ 120° . 
430 
4. CONCLUSION 
As the result to have tested the non-metric camera 
as collimator and to have applied this for observing the 
small size object more precisely, this researcher drew 
the following conclusion. 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B5. Vienna 1996 
  
This 
camera, | 
distortion, 
point of a 
the value 
calibration 
At the 
the result 
point was 
the difficul 
manufactu 
point. 
If one 
units, the 
coordinate: 
the numbe 
3-D coord 
If one 
140° |, the 
satisfactory 
angle inci 
proportiona 
remarkably 
So, it coul 
about 3-D 
ACKNOWI 
This = 
DIRECTEL 
Foundation 
Fioretti, 
Stereomet 
stereophot 
human 
Biomechar 
Fraser, C 
cameras i 
The Can 
259-279. 
Fryer J. ( 
Surveyors, 
pp. 330-3: 
Erver «J.. 0 
calibration 
Internation 
Remote S 
X X IX, Pi 
Hatzopoulc 
close-rang 
No.10, pp.