Ebner: 
  
measurements are expected to be ready. In principle 
also available inertial navigation systems could 
be used. Experimental results have been obtained 
in USA and Canada. Soon we shall have our own in 
Germany. As far as availability for practical 
application is concerned it can be expected that 
especially GPS receivers will become inexpensive 
enough to be applied in practice. Now, about the 
expected precision especially of GPS receivers: At 
the moment instantaneous accuracy in the 10-20 
meter range is obtained. It seems to be quite 
realistic, Fred Doyle mentioned it too, that the 
relative precision is better by a factor 10 or 
20. I was careful yesterday in my paper and quoted 
a factor of 5 - 10 only. For aerial triangulation 
a relative accuracy in position of a few meters is 
probably very realistic, based on pseudo range 
measurements. Phase measurements seem still to have 
problems with cycle slips. But I am sure the 
problem will be solved and that we can expect very 
Soon orientation data for camera position in the 
order of 1 meter or better. 
In the previous Session of Molenaar the question 
Of reliability of these additional data was 
raised. It is perfectly true that aerial triangu- 
lation will depend practically entirely on these 
additional navigation data, in the same way as we 
depend now on ground control. All the questions 
and problems related to reliability are basically 
the same. But I do think that the situation is 
really very much better with regard to orientation 
data because during an air flight mission we have 
continuous recordings of all these measurements at 
a very high rate, we speak of 10 or 20 or 50 Hz.- 
That means every tenth or 20th of a second we have 
a recording - perhaps not all the time, but at least 
in the surroundings when the camera is fired. So 
we have a lot of data to analyze before some data 
are introduced into the block adjustment as 
observed orientation data. I think with preproces- 
sing and preanalysis of observed orientation data 
we are basically in a much better situation than 
with ground control, as we have thousands of 
observations to analyse. Thus I do not consider 
the reliability of orientation. data to be a 
serious problem, although it will deserve attention. 
Thank you for this comprehensive and optimistic 
answer and I thank you, Mr. Friess, for your paper. 
46