SESSION 6: 
  
CLASSIFICATION AND SAMPLING 
  
  
CHAIRMAN: O. Jacobi (Denmark) 
AUTHOR: O. Jacobi (Denmark) 
TITLE: REPORT OF WORKING GROUP III/3 INCLUDING WORKSHOP IN 
OHIO, MARCH 14-15, 1986 
DISCUSSION: 
No discussion. 
AUTHORS: R. Bill and F. Steidler (Switzerland) 
TITLE: PROGRESSIVE SAMPLING USING TRIANGULAR INTERMESHING 
- AN EFFICIENT PHOTOGRAMMETRIC DTM DATA ACQUISITION 
METHOD 
DISCUSSION: 
  
Rauhala (USA): 
Bill: 
I'll just comment about the general character of 
this problem. Namely, once we start going from ana- 
lytical systems to digital systems we are getting 
an enormous amount of data. If we try to automate 
these systems, the sampling speed will be measured, 
say, in hundreds, maybe in the future even in 
thousands, of points per second. If we now compare 
the reported solution speed and the number of 
parameters we have a discrepancy of one thousand 
to one. Which means that in order to automate the 
DTM processing on-line in future digital systems, 
we need to have solution techniques that are at 
least one thousand times more efficient than the 
reported method. 
Ok, I agree with your comment. On an analytical 
plotter, where the operator sits in. front of the 
restitution system, it is a question of the reaction 
time of the system, but in on-line work it is the 
speed of such procedures that will increase very 
) much. 
61