ations and the Standarı 
ransfer and the azimuth 
point triangulation) 
ang 
cases and visua] 
ized in 
es Af and Mg, the other 
indicated in fig. 3a onl 
ty made in The Neth 
of 1600 m. 
r-tilt and y-tilt Were as. 
y, 
et 
of the ground along the 
besides the profile along 
of the chain of Thombs, 
"lines, arbitrarily chog 
' only nadir point trian- 
tical scale of the ground 
t of the diagrams 1.7, 
raphs, flying height and 
figure. The topography 
to caracterize the topo- 
it in hundreds of meter; 
parison — the standard 
are derived in Appendix 
n: 
rs Af and Aa, in scale 
s "g and m, in spatii 
an average topography 
ictor up to 1.0 they con- 
are somewhat larger. 
With standard (acciden- 
of accumulation of sys 
hat the really occurring 
nree times the standard 
nd azimuth transfer às 
1 even for hilly country 
continue to be generally 
tor of 3.0 (fig. 8 and 9). 
course the observation 
' a radial line is, on t 
line being greater. Tlis 
the angular coverage of 
so much the more suitel 
camera in the examples, 
| ih a standard error of 0.°75-1 
3 
+ 1-9. As a wide angle camera was chosen (fig. 1, 2, 4, 6, 8) a camera with focal length 
id 100 mm, picture size s — 180 X 180 mm?, implying a lens angle (across diagonal) 
d 104°— which, with a view to application for radial triangulation, is about equiva- 
: (to a camera, f = 6 inch, s = 9 X 9 inch2, a = 93°, Super wide angle cameras are 
Jen ptedly the answer for the mapping of vast areas of underdeveloped country [8]. 
pe examples (fig. 9, 5, 7, 9) was assumed a camera, f = 88 mm (3,5 inch), 
930 X 230 mm? (9 X 9 inch?), a — 123^, which is equivalent to a camera with 
lee lens [8], f = 70 mm, s — 180 X 180 mm?, a — 122*. 
Meanwhile it is evident from fig. 1-7 that the difference in accuracy of the transfer 
ıf scale or azimuth between wide angle and super wide angle cameras is hardly felt in 
he types of terrain assumed in these examples; it is easily understood that it must be 
suh more evident in more mountainous country ; see fig. 8 and 9. Nevertheless the 
apr wide angler is still to be preferred in view of the smaller number of rhombs in a 
dain of certain length. 
The flying height was chosen such as to produce pictures on a scale 1 : 60.000; this 
pans a flying height of 6000 m for wide angle cameras or 5280 m for super wide angle 
meras. An exception is fig. 1 where the flying height was assumed to be 4000 m to 
I roduce a picture scale 1 : 40.000, à scale which has been widely used in the past for 
mall scale mapping. The great improvement in recent years of the o 
amera lenses however, tends to substitute this scale by the more econo 
it1:50.000 to 1 : 60.000. 
The x- and y-tilts ¢ and «o, represented in fig. 3a (with which correspond an 
werage total tilt à = 33’ and a maximum of 54’) and referring to a routine photo-flight 
iy the K.L.M., were obtained from a spatial triangulation of the strip. These values 
my be considered representative for the precision in photo-flying obtainable nowadays 
i altitudes of 4000 of 6000 m provided that an appropriate aircraft with auto-pilot is 
md, while the crew is experienced in photo-flying and especially instructed about the 
importance of small tilts. It is a matter of fact that present day air photography only 
aeptionally shows tilt values of more than 1° [9]. From the numerical examples 
lustrated in fig. 1-7 it is evident that this is quite adequate for principal point trian- 
mlation of flat or hilly country. 
ptieal quality of 
mic smaller scales 
For more mountainous country or when higher accuracy is required, nadir point 
mangulation is to be used. 
To locate the nadir point on the photograph the cameratilt must be known. There 
ue various methods of tilt determination; methods which us 
ie flight and methods which only use the pictures as such. 
One of the earliest attempts to determine the camera tilt was made in Finland by 
lie use of horizon cameras. This method, still used in that country, yields x- and y-tilts 
.25 only. [10]. Some other countries consider also the 
Wmduetion of this method and it is noteworthy that in The Netherlands a horizon 
nra was developed which photographs the entire horizon. 
À method which seems to be of more universal use — it being independent of the visi- 
ility of the horizon — is gyro-stabilisation of the camera or gyro-recording of the vertical 
ining the flight. The latter method probably is apt to give a higher accuracy although 
hreent testflights made in England, using a gyro-stabilized camera, the camera-axis 
id not depart from the vertical more than + 15 minutes. [11]. An American firm adver- 
e special apparatus during 
| ling an optical gyro-stabilized camera-mount claims even an accuracy of 6. [12]. A 
qo device for recording the vertical has been used in practice already several years in 
Trance and the accuracy obtained, using three gyroscopes simultaneously, is 7' - 14' [13]. 
À device for the Same purpose has been developed in Italy and is under test now. [14]. 
^diminary results show that the nadir point is determined with an accuracy of 5’. [15]. 
À different approach to the problem has lead to the construction of the solar peris- 
Me [16] and a celestial tilt indicator [17], both of which instruments record the nadir