——— 
  
Consequently, relative accuracy is maintained to a fairly high degree and for 
this reason it may be expected that by adjusting the block to a few points, all the 
neighbouring points are also determined with the required accuracy. If, on the 
other hand, the standards are of a higher order so that the relative accuracy over 
larger distances is not adequate to produce the desired result, then the distribution 
of control should be denser to prevent the remaining errors in between the control 
from becoming too large. 
CONCLUSIONS 
'The method as discussed in the previous paragraphs has been developed in an 
attempt to improve upon other procedures of single strip adjustment. The result 
is achieved at substantially the same cost compared to the classical methods of 
graphical transformation of each strip. The only extra work involved is the deter- 
mination of the parameters a, B, y, etc., for the connection of strips along their 
entire length. Because the coefficients of the equations are specially chosen, and 
also because the same coefficients apply to all runs, the computations assume a 
routine nature, and require very little time. Moreover, the extra time spent in this 
stage is partly compensated by the fact that less attention needs to be paid to the 
initial scale setting of each run during the bridging. Any linear scale difference 
between the runs is eliminated by the linear transformation or connection of runs 
in the first stage of the computations. A definite advantage of the method is that a 
substantial saving of control and field cost is obtained. 
It is difficult to postulate the minimum control requirements for single strip 
adjustment as the matter is rather controversial. With a bare minimum of three 
points on each side of a run and so arranged that they are common between adjacent 
runs, 12 X 3 = 36 points are required for 11 runs. If this can be reduced to nine 
points, as previously suggested, it means an actual saving of 75 per cent will be 
effected. 
The method appears to be particularly useful for topographic mapping of areas 
where it is difficult or costly to put in a dense pattern of control. 
Test Area Data 
An area of approximately 1,000 square miles covered by 11 runs with an 
average of 18 photos each, was tested by adjusting it to 9 control points arranged 
as in Fig. 2. 
Photography with RC5 camera, F = 115 mm. 
Flying height approximately 15,000' above ground level = 4,600 metres. 
Picture scale approximately 1 : 40,000. 
Aerial triangulation in A-5 at approximate scale 1: 15,000. 
120 regularly distributed points were established by ground survey, of which 
nine points were used in the adjustment. The mean square errors and frequency 
distributions pertain to the remaining 111 check points. 
Residuals ın X at 1:15,000: 
Error Frequency Relative accumulative frequency 
mm. To mm. 
0 31 28—0 
0-1 44 67 «01 
0-2 22 87 < 0-2 
0-3 11 97 < 0-3 
0-4 1 08 < 04 
0:5 2 100 < 0-5 
my, = 0:16 mm. at 1: 15,000 
— 0:06 mm. at picture scale 
z- 2:6 yards on the ground