owards a cop. 
© ratio between 
‚Ween distances 
te the effect of 
ling dy by the 
S. 
ive map posi- 
ohotograph can 
obtained from 
» only a single 
o long distance 
ngulation strip 
* to fulfill this 
established, the 
comparison of 
ates. This com- 
ro at both ends 
the transversal 
ors. Therefore, 
determined by 
. Therefore, an 
ry. If the con- 
order to com- 
À detailed con- 
yublication now 
been used for 
aft from the 
plenty of scope 
y designed for 
) far. A longer 
oblique photo- 
I] under study. 
| of the method 
e, and the con- 
proportional to 
o 
etermined with 
1e final errors. 
as possible. In 
eans it follows 
n the contrary, 
from the nadir 
e for the bend. 
rhout the strip 
5 
ered py vertical photographs, the method can not be applied without considerable 
lification. 
| The construction of the straight line on the diapositives is the most delicate step in 
m procedure. In past experiments a calibrated steel straight edge has been used to guide 
pecially designed knife. The line produced by the knife is very fine: it is har dly visible 
he naked eye. The main difficulties are in identification of the transfer points on two 
ceding oblique photographs and in orienting the straight edge. However, these dif- 
ulties may be reduced in the near future, since a special “Straight Line Plotter” is 
wing built for the line construction. This instrument will be equipped with an optical 
| stem enabling simultaneous viewing of two oblique photographs. Two photographs can 
| shifted separately in the horizontal plane in two directions, x and y, and rotated about 
vatical axis. Both will also be moveable together in the y-direction. The first mentioned 
gions provide means for ‘relative orientation” and the common y-motion represents a 
might line. 
  
st Results. 
The method is being put into operational use by Topographical Survey of Canada. 
Inst experimental tests were made over the Canadian U.S. boundary. Total length of 
"strip was about 240 km. Approximately one-half of the strip covers hilly terrain with 
ferences of elevation up to 600 m. A large portion of the area is covered by forests. 
Photography was made from 6000 m using Fairchild K 224 camera for vertical 
\ntographs and a 6" Williamson camera for oblique photographs. Angle of inclination 
"Ihe oblique camera was 60° from the vertical in the aft direction. Kodak infrared film 
ith the proper filter was used for oblique photographs. 
| Bridging was done on a stereoplanigraph C8 and the bend of the strip adjusted by 
ng straight line points. Result was checked against 81 geodetic boundary moments. 
In straight lines were constructed. Because the flight was not very straight, only one 
(them could be carried through the entire length of the strip, the other covering only 
m one half of the length. Also each half of the long line was analyzed separately, and 
results are tabulated below. 
  
  
  
  
TABLE I 
| | Accuracy in Determination of a Straight Line 
| linc | Line Ground | Numerical Construction | Graphical | construction 
| No | Length Elevation i of a saigne line A of a sir sieht Heri 
| Differences | Mean | Max | Mean | Max 
| | y Error Error | Error Error 
| | | 
| 1 120km | 600m | 35m + 79m | 
[1 | 120km 600m | | 52m. | —18.7m 
| 2A 120 km 300 m | + 4.6m +14.9 m 
| 2B 120km | 600m | | =10.0m +15.9 m 
|? 240km | 60m | 94m +15.3m | 
  
  
  
{ 
| 
| Deviations in the case of the 240 km line are not all random but a very distinct shift 
3pesent at a particular point. At this point the control line was close to the edge of 
"strip. At the same time the terrain changes from flat to hilly. This may explain the 
"Wrence of a systematic error.