the 
th 
on. 
Or- 
not 
On, 
not 
cal 
fa 
ore 
Te, 
ion 
er, 
led 
m- 
her 
the 
re, 
ore 
to 
or 
vey 
m- 
Ve 
1ps 
ng, 
ct, 
TiC 
ich 
ips 
ter 
  
GRAPHICAL OR NUMERICAL PHOTOGRAMMETRY?, DISCUSSION 133 
maps have been finished. 
Taking for granted that photogrammetry is 
the indicated tool for the production of maps, 
the process, of which mapping is a part, fixes 
more or less the time limits within which the 
results have to be produced, and the specifica- 
tions, such as the scale, which have to be ful- 
filled. In this respect, it is not sufficient to judge 
the merits of the graphical or numerical photo- 
grammetry by comparing them in a limited 
sense, but it is necessary to look for their ad- 
vantages and disadvantages in relation to their 
position in the total project. This may result 
in the conclusion that the difference in cost 
is not important with respect to the saving 
in time which can be obtained. It may also 
indicate that somewhere in the process results 
may be more favourably given in graphical 
form, in other cases in numerical form. 
If the numerical method is chosen, an addi- 
tional point of co-ordination and co-operation 
is the question of how far and in which way the 
photogrammetry can meet the requirements of 
the project director. Here the data processing 
expert must be included in the team of people 
who organise the whole project. 
It is obvious that the application of numer- 
ical photogrammetry is closely connected with 
the automation of the procedure. The presented 
papers show clearly that great progress has been 
made, and that data processing is already a 
standard procedure in many cases. 
Just as in the case of cadastral applications, 
| personally am convinced that the correct 
solution depends to a great extent on local 
circumstances, and that it would not be correct 
to draw general conclusions from the experi- 
ences which have so far been published. In 
several cases, for instance in the digital deter- 
mination of profiles, special instruments are 
used, but those instruments are not yet avail- 
able on the market, and their output cannot be 
judged accurately so far. 
Without going into more detail on this 
subject, I feel there is a need for more informa- 
tion on cost, the capacity of the instruments, as 
well as a knowledge of the limiting factors of 
the system and the best organisation. I think 
Commission IV has rendered a great service to 
the development of photogrammetry and will 
continue to do so if it provides us with as much 
specific data as possible. 
To conclude, I wish to express my gratitude 
for the great help and the large number of 
suggestions I have received from my many col- 
leagues whom Dr. Härry has already named, 
without which I should not have been able to 
introduce this subject here. 
Discussion 
Dr H. HArry: We must thank Mr Van der 
Weele for his paper, and Mr Van der Weele will 
himself lead our discussion which will now 
follow. 
Mr A. J. VAN DER WEELE: I would now in- 
vite people from the floor and from the panel 
here on the platform to give their views on the 
subject. If there is anyone who has a question to 
put or remarks to make the people on the plat- 
form here are available to give, if possible, the 
right answers, so I should be very glad to have 
your reactions. 
Mr S. G. MOLLER: In our work in Sweden 
concerning re-allotment and large-scale map- 
ping, we have found that gross errors can more 
easily be detected by graphical methods than 
by numerical. We have many troubles just con- 
cerning the possibility of eliminating gross 
errors from our calculations. 
Mr A. J. vaN DER WEELE: Thank you, Mr 
Moller, for your intervention. I think the answer 
to this question could best be given by one of 
those who has experience of this subject. What 
about Mr Zarzycki? 
Dr J. M. ZARZYCKI: My particular expe- 
rience on numerical photogrammetry is an ap- 
plication to engineering projects, mostly to 
highways and railway design. Here a graphical 
solution is economically not feasible, so in 
applying numerical data we have à number of 
self-checking methods which pick up data which 
are in error. However, we have to take into 
account the fact that in having numerical data 
it is not so easy to recover errors, but this 
method is much faster, and the fact that it is 
much faster and is capable of handling a great 
amount of data is in favour of using numerical 
data instead of graphical. 
Mr W. T. PRYOR: It is a pleasure to have the 
opportunity to come here to London to partic- 
ipate in the 9th International Congress of Photo- 
grammetry, and particularly to comment on the 
fine paper by Mr Van der Weele. Needless to 
say, I need not repeat the many distinct factors 
which have an effect on the accuracy and utility