0 0 0 0 0 0 
U-— = | O 0 0 0 0 0 (4. 135) 
xx 1/2 0 0 1/R 0 0 
for estimting the coordinate changes being due to the matriz 
à and the error in (4.8) made by using (8.3) instead of (4.9) 
we set À = E and calculte some examples (Table 4.1). 
  
  
Table 4.1 
Agsunt of remainder terms (matrix R) and their errors 
dx=dy= Smm dr=dy=10nn x=dy=20mm 
my c x! wy! 2 Bye AR iy AR 
mn Lum um um um uri um un 
90.000 590 100 -1.4 *0.1 -—5. 7 *1.1 
A a CSS 407 
900.000 300 400 -1.0 0.0 -4, 2 +0.4 | —10.7 *3.3 
a et dorée: ais eee 20. 200 0:0 0.407 40.4 335 3,3 
Table 4.1 allows the following conclusions: 
— 
* 
Choicing Aap = Arne = 10mm, it is possible to set the 
martix R = 0 in (4.8) for the restitution of aerial photo- 
graphs. In (4.3) the influence of the earth's curvature is 
adequately taken into account by the integration of the 
first part of the remainder term in (4.5). 
2. For large format Space photographs R must be taken into 
account o» dA > Tue Dust De reduced to 2mm. 
3. Vhen changing from one segment to the other errors of 
2:4R occur for R + 0. For da e o dye X 10mm these 
differences lie below the measuring accuracy of the 
equipment and of the photographs. 
4.4.2.2. Other applications 
In Section 4.4.2.%. the eartn's curvature was singled out in 
order to demonstrate the Principle of the algorithm by way 
of an example. In the practical realization of a program 
curvature of the earth and refraction will of course pe taken 
into account in common. Since their effects are opposite, the 
residual errors shon in Table 4.1 are further reduced. 
Other application possibilities exist in architecture and in- 
dustrial photogrammetry in the restitution of cylinders with 
circular or elliptical cross Section. For a cylinder with 
circular cross section (4.1) becomes (Fig. 4.3) 
X= xz, *r + E)sin(2/r) 
7 y 
Z = Zu — (r * Z)cos(2/r) 
Problems of two-media photogrammetry can likewise be solved 
in this way. 
- 510 -