28
Î7 feet (Î2.2 metres) respectively. The above results demonstrate that the
horizon camera provided more accurate results than the statoscope.
7.2.3 Accuracy of Aerial Triangulation
To evaluate the accuracy of aerial triangulation over long
distances using the horizon camera data, a block of 8 strips was chosen. Each
strip was 130 kms long and comprised an average of 27 stereo models. The
photography was flown with a Wild RC-9 camera at 1:50,000. Fig. 15 illustrates
the actual errors in elevation observed in Strip 5 before any adjustment was
made. They are typical of strips bridged using horizon data. From this figure
it can be seen that errors in elevation propagate linearly. After bridging a
length of 130 kms. (27 stereo models), the closing error is only 61 feet
(19 metres). The closing errors in other strips were of the same order of
magnitude.
Adjustment of all eight strips and of the block was carried out
utilizing vertical control only at the beginning and at the end of each strip.
A sketch of the area showing the layout of the flight lines and location of
ground control is shown on Fig. 19.
To determine the residual errors in elevation after block adjustment,
So
known elevations of "24 points located within the block were compared with the
elevations established photogrammetrically. The results are:
Ah in
feet
Ah in
metres
%
0 -
t 5
0
- tl.5
50
t 5 -
iio
tl.5
- Î3.1
30
iio -
tl5
Î3.1
- Î4.6
10
Î15 -
Î20
+4.6
- Î6.2
10
The mean square error in elevation observed on these points is m^ = t9 feet,
or approximately -3 metres.
7.3 Accuracy of Doppler Distance Measurements
To evaluate the accuracy of the scale determined from Doppler
navigator data, we compared the base distance given by Doppler with the base