Table 1: Comparison of different DTM-acquisition methods 
  
  
  
method measuring time number of points accuracy 
(h: m: s) (-) (m) 
Prior information 2:09:00 6655 - 
40 m Profiling 0:34:45 702 1.43 
40 m Grid 0:21:20 590 1.49 
10 m Grid 3:20:15 9118 0.87 
Contourlines 1:57:00 9290 - 
PSA without 2:46:15 1577 0.87 
prior infos 
(3.densification) 
PSA with all 2:28:25 1065 0.90 
prior infos 
(1.densification) 
PSA with subset 2:29:50 1289 1.13 
of prior infos 
(2.densification) 
  
  
  
  
  
  
  
In the first variant of PSA three densification steps were made. A 
comparison of plots of the morphologic information with the point 
distribution in the PSA acquisition without these information shows 
that PSA densified exactly in those regions where the terrain is 
structured. Normally it would not be necessary to capture prior in- 
formation, or not as dense as would be necessary for the grid mea- 
surement. So the time is always reduced by more than 20 % compared 
to the grid. The number of registered points is only 20 % of that 
which would be necessary for the grid measurement. And the quality of 
the DTM is minimum the same as with a dense grid. 
These values are also valid if one compares the grid measurements 
with the Progressive Sampling with prior information (2. variant) on 
the terrain. Including all 4446 points which is too much for that 
terrain allowed one densification step in PSA in 75 % of the time 
required for the grid measurement. The number of points was reduced 
up to nearly 90 %. The third variant shows a compromise solution, 
where only a subset of terrain characteristics 1ike ridges, scarps 
etc. were included. Here the terrain analysis did not take such a 
long time and one more densification step could be done in the same 
time. 
- 76 -