me units as 
—days, 
sis of work 
on routine. 
tegories 
al modeis 
en the 
or at least 
project area, 
with differing 
process- 
the process 
Se 
ess components 
a 1038 of 
nimetric 
er (the remain-- 
rs of subsequent 
pective 
wever, due to 
ht often be the 
A5, — 
By far the most complicated task is the establishment of accuracy models for the 
determination of minor control. The reasons for this are the large number of 
influencing factors and possible combinations between various preparation — 
observation and adjustment methods, and the influence of the large number of 
possible configurations and flight patterns of the photographic coverage. 
Therefore accuracy models should be established, considering only the most 
important process parameters as variables and considering other factors 
(especially those pertaining to the executing organisations) by applying 
calibration factors, which can be assumed to be more or less constant for an 
individual organisation. 
Three different methods can be used for the establishment of accuracy models: 
a) Analysis of practical experiments. 
b) Theoretical analysis using functional and stochastic models of the actual 
‘physical process 
c) Application of computer simulation 
To the author's opinion the last method in conjunction with carefully designed 
practical experiments offers the best results as more realistic assumptions concern- 
ing the error sources and their stochastic properties can be applied as in method 
by. 
21, Establishment of semantic information models 
  
This information is required by the planner in order to make sure that the semantic 
information (i.e. the types and kinds of natural and cultural features) required 
by the product specification can be achieved. 
In the first place this will depend on whether these natural and cultural features 
can be identified and classified on the aerial photography with sufficient 
reliability. This is primarily a function of the photoscale, the type of camera 
and film-filter combination used and the photographic processing. 
Furthermore we have to consider the loss of information due to the various 
processes applied, the optical properties of equipment used, the Skill of inter- 
preters etc. 
Alternative means for the planner to achieve the necessary semantic information 
include the application of premarking (signalisation), field identification 
carried out before plotting and field completion. Therefore, from the economic 
standpoint there exists a very strong trade-off position between on the one hand 
the choice of photoscale and the cost of resulting photogrammetric work and 
-