Engineer Colonel The Ninth International 
Karl G. Läfsträm Congress of Photogrammetry 
Hq. Staff of the Defence Forces Committee | 
Helsinki September 1960 
HORIZON CONTROLLED AIR PHOTOGRAPHY 
As mentioned in the invited paper of Commission I," Survey Navigation and Determination 
of Camera Orientation Elements", by F. L. Corten ( 4.3 page 271}, Wild has constructed 
a modern Horizon Camera for determination of the angular orientation, which records 
four horizon images synchronized with the main camera. Both cameras are stabely mounted 
on a mutual camera platform thus, that the axes of the four horizon lenses are at right 
angles ro the optical axis and parallel with the image coordinate axes of the main camera. 
This new horizon camera coupled with RC8 has quite recently been tested in Finland and 
the preliminary results are now available. 
Test material 
- One strip (flight from East to West ) of 19 pictures { NR 613-632 ) in 1:30 000 ( c = 
15. em, h- 4500 m, overlap 6096 ) of a normal survey flight in the North of Finland. 
-  Signalized geodetic points of control ( x,y,z } in distances of about 10 km for scale 
determination and leveled ( not signalized ) points of vertical control along the edges 
of the strip for separate absolute orientation of every model in a stereo plotting 
instrument. 
- Horizon images taken on 36 mm Panchromatic Film ( Agfa Isopan ISS 21 ) using Infra 
Filter ( 670 mp ). The images were almost cloudless. The front ( West ) and left 
(South) horizon lines were clear and distinct, the right ( North ) was — somewhat 
hazy, but stil measurable, whereas the rear { East } horizon was so dim that no 
refence line could be found. 
Determination of the angular orientation of the camera using Horizon Images was performed 
as has been published / 1 / as well by monocular as stereoscopical measurement. In order 
to get a greater accuracy the relative inclinations of consecutive photographs were in 
both cases calculated from the vertical displacement and not from the tilt of 
the horizon line. 
  
No special measuring devices for this small size horizon pictures { 9 x 24 mm ) have jet 
been constructed. Thus a normal protractor (diameter 95 mm) and a Zeiss^ stereocomparator 
had to be used. 
For the monocular measurement the small horizon pictures were enlarged ( 4.3 times ) on 
film and used as diapositives. The accuracy obtained with the protractor is of the same 
order as when the earlier big size horizon pictures were used. 
The front horizon images were measured and the relative f? - values ( Pitch) were cal- 
culated separately twice. The standard error of the mean value of these two measurements 
proved to be + 2.0% {0.029}. 
The relative co - inclinations { Roll Juels denermined separately from the left and right 
horizon image. The mean value of this two independent measurements showed as well a 
standard error of + 2.0%, 
For the stereoscopical measurement the midmost horizon image ( NR 621 ) of the strip was | 
chosen as reference image and as original negative placed in the left frame of the stereo