164 DISCUSSION ON THE GENERAL REPORT OF SUB-COMM IV-3 
certainly think that with such tools they could 
have been far better than the diapositives we 
had to work with. 
Due to the time factor we limited the evalua- 
tion in the instruments to three different models, 
twice in each model. That was done so that the 
Hydrographic Department with the A7 Auto- 
graph of the Institute of Technology could treat 
(a) and (b), the National Road Board treated (a) 
with an A8, and the Geographic Survey treated 
all three models with an A7 automatic punch 
tape recording of the co-ordinates. 
There was a little variation in the number of 
signals which were pointed by each operator due 
to interpretation difficulties. In each treatment 
every point was pointed twice, and the average 
was used. 
The instrument co-ordinates were trans- 
formed with a Helmert transformation with an 
electronic computer, two ways. We first made it 
on all the points; up to about 100 transformation 
points were used. The second time it was only 
on 10 well distributed points, and it was about 
the same result. We took as many as 10 because 
we did not like to introduce any errors which 
might be in one or other geodetic points. 
We were faced with a final error which was 
not of the same magnitude but only about three 
times as big as the errors in the signals — the 
mean error — so we had to be careful. The mean 
of all six models was the mean square value of 
the discrepancies between the photogrammetric 
and geodetic co-ordinates, 25 mm in x and 
24 mm in y; in radial it was 34 mm. 
That was when transformed on all the points. 
Those values are about exactly the same, as a 
matter of fact they are a little lower. When we 
transformed only ten points there was no signi- 
ficant difference. 
Of course, it was not only a matter of x- and 
y-co-ordinates. There was also the z-co-ordi- 
nate. But we had no problem for the electronic 
computer to adjust the model to z, and so we 
made it graphically and rather simply and 
found the average error to be about the same. 
The standard error was about the same, that is 
about 30 mm. 
The conclusions are mentioned in the short 
report which has been distributed to the Con- 
gress. The title of the report is “Precision Aerial 
Photogrammetry from Very Low Altitudes”. 
This is offered by Mr Thunberg and myself. We 
say: 
“It is quite possible to take aerial photo- 
graphs from a normal speed aircraft from 
very low altitudes under almost any weather 
conditions at any time of the year and deter- 
mine co-ordinates of signal points with an 
accuracy in x, y and z of about == 30 mm." 
I might add there that one idea we have is 
that the accuracy must be such that construction 
people — for instance, in harbour — can use 
such signal points if they are marked well on the 
ground; they can use them as reference points 
for building work, for construction work. I 
might also add before I go on with the conclu- 
sions — as I am not sure I mentioned this — 
that the sun was about 13^ above the horizon. 
I was very much interested in the slides which 
Professor Schwidefsky showed. That does not 
give me much light, but it is certainly enough. 
“Plotting of details and control lines for very 
large-scale maps is possible even in under- 
exposed negatives from this very low altitude, 
if the ground surface contains such contrast 
and texture that the greyness of the film will 
not cause an erroneous stereoscopic impres- 
sion. In the described test, this was, however, 
on the lawns of a park." 
On the grass lawns, however, there was too 
little difference to put the floating mark on the 
surface. 
That is what I have to say. 
Dr H. HArRry: I would like to thank Mr 
Fagerholm for his communication. My opinion 
is that he has achieved very fine results and has 
given us a good look at our work for the future. 
Are there any other people who have experience 
in this matter? 
Mr. H. G. Dawe: Dr Schwidefsky and Mr 
Fagerholm have, in fact, covered most of what I 
had in mind to say, but I should like to add a 
comment or two. Firstly, I should like to say 
that our experience in the use of photography 
taken under adverse weather conditions was 
born out of dire necessity. Here in this country 
there are very many days — consecutive days — 
when the cloud basis sits at about 2,000 to 
3,000 feet, days where the light is not bad at 
all, sufficient to register on the light meter, and 
with the super-fast films we have certainly ca- 
pable of producing a negative of sufficient 
strength. 
There is one danger, however, in that when 
one has opened the possibilities of photograph- 
ing under adverse conditions, one wants to go 
on photographing under worse and worse con- 
ditions. We have found that our air crews and 
photographers tend to take photographs under 
conditions which result in too long a develop- 
ment and a build-up of grain. These super-fast 
films work extremely well providing they are 
nol 
the 
pre 
L.C 
poi 
she 
pic 
res 
ple 
ael 
pa 
vel 
tic 
pr 
sol 
ra 
ye: 
ing 
alr 
mz 
ph 
soi 
sui 
yo 
tol 
sal 
qu 
pa 
pe 
lor 
éc 
nu 
vo 
alt 
ne 
de 
pa 
nu