transformation. The derived transformation parameters are then capable of 
relating the new physical situation with the old one. The logical structure 
of the operation is sketched in the flowchart of Appendix A-2. The restored 
¢ 
optical and numerical models are identifical with the previous ones within 
the range of a few micrometres, as characterized by discrepancies in the 
transformation fit on fiducial marks after the new inner orientation. The 
new model can be checked and remeasurements carried out as needed. In most 
instances, the misidentification is corrected and new bridging of the 
remaining photographs in the strip is not necessary. 
CONCLUSIONS 
By choosing tie points during the measurement process, one avoids problems 
otherwise associated with their identification and drastically reduces the 
time necessary for preparatory triangulation work. The tedious and time 
consuming selection of the points in image overlaps becomes unnecessary. 
Also redundant is the point transfer by mechanical marking of photographs in 
special devices and manual point numbering. The NRC on-line triangulation 
represents a fast and efficient procedure with an adequate checking power to 
produce accurate data praeticaliy free of gross errors and biunders. Thus, 
it greatly simplifies the ultimate off-line block adjustment of data. 
REFERENCES 
Jaksic, Z., 1979, Software Design and Basic Application Software of the 
Anaplot, The Canadian Surveyor, 33(2): 131-152. 
Kratky, V., 1979, On-Line Analytical Triangulation, The Canadian Surveyor, 
33(2): 162-175. 
Kratky, V., 1980, On-Line Aerial Triangulation on the Anaplot, Proceedings o 
The ASP Analytical Plotter Symposium, Reston, pp. 207-224. 
iti 
Kratky, V., El-Hakim, S.F., 1983, Quality Control for NRC On-Line 
Iriangulation, Photogrammetric Engineering and Remote Sensing, 49(6): 
763—769. 
      
  
   
    
   
      
  
    
     
    
   
    
   
   
    
    
APPENDIX A-1-2 Operator svstem interaction in NRC triangulation