EEE SETS 
ARE RL 
NETS 
e D 
LE. 
— 
Roues 
se . 
d 
= 
= 
Ds 
idle 
  
  
  
  
  
  
  
  
  
       
        
  
ti 
  
  
  
ec 
  
    
    
   
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
   
   
        
Dh 
IT. In the second method a universal measuring machine, SIP MU- 
214 B, Nr 277, was used. The instrument belongs to the Division of 
Mechanical Technology at the Royal Institute of Technology. The ca- 
pacity of the machine is 10 7X 40 cm, which made it necessary to mea- 
sure the grid in sections, 10 X 24 cm. Four sections were used, so that 
the central 20 X 20 cm of the grid were covered twice. In each sec- 
tion ten points were measured, and in two of the sections some more 
points. 
The coordinates for each section were obtained as the mean of six 
observations. Their standard error, as calculated from the differences 
between several observations at the same point, was estimated to be .37 
microns. 
The first step in the calculations of grid coordinates was a coordi- 
nate transformation, which was made to bring the measurements into a 
common system. The angle between the two slides of the instrument 
was proved to be skew, and this skewness had to be taken in account. 
The adjustment computations of this step became rather comprehen- 
sive. 28 unknown quantities were to be determined from 48 observation 
equations. The standard error of an observed coordinate was calculated 
to ./0 microns. The skewness of the slide movements was 3.24 . 10— 
radians, with a standard error of . 34 . 10—? radians. 
In the second step, the coordinates, determined as unknown quanti- 
ties in the first step, were compared with the given coordinates by 
means of an orthogonal transformation, including a small change of 
scale (— 2.03 . 10—?). The standard deviation of a coordinates was 
estimated to .85 microns in this comparation. By combination of the two 
sets of transformation data, it 1s possible to compute coordinates in the 
given system from the measured coordinates. 
The result of the two measurements of the grid show, that the coor- 
dinates given by Wild Co., very accurately correspond to the true po- 
sitions of the grid points. 
The rigorous computation of the first step in the coordinate trans- 
formations proved to be very time-consuming. It is of great interest to 
find easy methods for the connection of several different sets of coor- 
dinate measurements. 
Reference 
[1] Hallert, B.: Tests of Fundamental Photogrammetric Operations. The Photo- 
grammetric Record. April 1955.