36 mm 
+ No.4 
d 
| No.5 
T: + c = a + [24 mm 
No.3 | No.1 
10mm b 
| 
dc + No.2 
| 15mm | a: SIDEI e:SIDE M 
b: SIDE Il d: SIDE IV 
Fig.1 Fiducial Mark of Glass Plate. 
EE neoctive mask 
[m9 799 — -- 7] positive mask 
light source 
  
  
vacum pump 
_— Base glass plate 
glass plate, > mas plate 
photo emulsion 
Ü 
exposed area 
[er Cr oiching 
  
me mark J 
fiducial glass plate 
Fig.2 The Process of Cr Etching for Glass Plate. 
    
   
   
     
  
  
  
   
  
   
  
  
  
   
   
   
   
   
   
  
     
  
  
   
  
measuring the difference between No.5 fiducial 
mark and the center of central slit of collimator. 
Radial lens distortion coefficients : Total 18 slits 
were imaged as shifting and rotating the single 
collimator with 6° incresement from 0° to +24° 
successively to horizontal direction, with 4° 
incresement from 0° to +16° to the vertical 
direction for obtaining the radial lens distortion 
coefficients. Collimator rotation angle (minimum 
1/1000sec) was controlled by computer. And then 
we measured the distance between slits on the 
developed negative film (Table 1). 
There being considered distortion free around 
center of lens at range of +6° in horizontal and 
+4° in vertical, EFL(Equivalent Focal Length) was 
computed as 36.266mm in case of +4° among 
exposed slits. (radial distance / tan 6°) 
Theoretical values of radial distance to slits,which 
have to be imaged from collimator, were 
deducted from the calculated EFL and the values 
were compared with the radial distance to slits 
imaged actually and the  calaulated radial 
distortion is shown in Table 2. : 
Radial distortion coefficients in each side were 
obtained by Brown polynomial based on distortion 
data taken from Table 2 (Brown,1971). 
CFL : With the CFL(Calibrated Focal Length), 
we took an average on the amount of horizontal 
distortion on side I and III by the formula ( ri- 
CFL * tan 6i) + ( rj - CFL * tan 6j) = O and 
maximum and minimum values. We obtained a 
36.079mm average value from 36.131mm and 
36.027mm on vertical direction side II and IV . 
The amounts of distortion in horizontal dir- 
ection range from 0.009 to -0.144mm in case of 
application of EFL while range from 0.004 to 
-0.051mm in case of application of CFL and then 
Table 1 Collimator shift angle versus radial distance. (Direction to horizontal 
& vertical) 
  
angle 424? . 418? 412? 
H 
+6° 0° -6° -12° .18- -24° 
  
dist. -16.003 -11.718 -7.696 -3.810 0 3.821 7.704 11.735 16.020 
  
angle +16° +12° +8° 
pan pete 0 ge Page 7 4 
  
V 
dist. -10.356 -7.691 -5.086 
-2.5310 2.541 5.093 7.684 10.346 
  
CALIBRATION TEST 
Collimator method 
PPA and PPS : Setting up the fiducial glass plate 
on camera mount, We found out PPA (Ax -103pm, 
Ay = 30pm) and PPS(AX = 7pm, y = Opm) through 
decrease by 64% than in case of EFL. The 
amounts of distortion in vertical direction range 
from 0.005mm to -0.053mm in case of application 
of EFL while range from 0.019 to -0.006mm in 
case of application of CFL and then decrease by 
76% than in case of EFL. Consequently we 
recognized that CFL is to be applicated.