shown that the idealisation "errors" of these boundaries in an aerial survey
are of the same order as those in a field survey. Expressed in errors in
distances, relative to reference points, the standard errors due to idealisation
may be of the order of: 2—10 cms for walls, 5—25 cms for hedges, 10—40
cms for ditches, etc. Now, it would be the height of folly to crawl under
a hedge, materialise the idealisation of the hedge with wooden pegs and
next to measure the positions of these wooden pegs, relative to the
measuring frame work, with one centimetre precision ! It is clear that a
repetition of the idealisation would produce positions of wooden pegs
which might have random shifts of several centimetres (or decimetres,
depending on the width and the symmetry of the hedge), in the direction
perpendicular to that of the boundary itself, as compared with the first
series of pegs. An economic proposition for the measuring accuracy of a
boundary survey is to require that this should be no better than the accu-
racy of idealisation !
Assume that we have an area where the cadastral boundaries are
mostly features with ( 0 à ) idea] 7 20 eM.- If numerical stereo-restitution is
considered, using an analogue instrument, the distance measuring accuracy
of this procedure may be assumed to be (vide the results of Reichenbach):
(04 )photogram < 17.5 microns (a conservative estimate). Accepting a
1:1 ratio between measuring accuracy and accuracy of idealisation, the
photo scale to be used can now be computed from
17.5 u = 20cm +1: 11,500
and the final accuracy of the distances is then 04 7 V 20° + 20° = 28 cm.
1.4.2 Graphical Stereo-Restitution
If, in the case above, graphical stereo-restitution is considered instead
of numerical stereo-restitution (because the former is simpler and cheaper),
then accuracy is lost due to graphical plotting (04 7 0.15 mm at the scale
of the map) and due to the scaling-off of the distances from the map
(0g = 0.10 mm at map scale).
The final accuracy of the distances, in the case of a map scale of
1:2,500,is then