" 13 
difficult to change by the user the Fortran coded 
orientation routines may be changed at will if the 
exchange data format between the main program and the 
orientation routine in the common data area is maintained. 
1. The relative orientation program in use stores the 
coordinates of the transfer points so specified. 
2. For these it calculates photo coor dinates and compares 
them with those calculated from the point measured to 
<? , ,V' • - • , 
obtain px, py. 
3. The px for the transfer points is used to scale the model, 
changing Abx. 
4. The py values are used for an adjustment of relative 
orientation by least squares. 
5. The coefficients for the A matrix are formed by differences 
obtained in calculating photo coordinates with the 
respective rotational elements changed by o,l° and the 
translational elements by 1 IRm * 
6. Thereafter the orientation elements are corrected, 
7. The program is iterated until the residuals are less 
than 10 ; iXm or unless a program switch is depressed. 
The present absolute orientation program utilizes an 
iteration procedure: 
1.It first assumes to be equal to zero. 
2. Then it calculates the 3 independent coefficients A, B, C 
of the orthgonal rotational matrix R as well as the scale 
factor iteratively until changes become small, 
3. Then the transformations to a levelled model y^, z T 
ape applied. 
4. Then linear transformations are carried out in the plane 
changing z with ^ only.