(17)
(18)
(19a)
(19b)
(190)
(19a)
(19e)
(19f)
(20a)
(20b)
(20c)
(21a)-(21c)
In the same way as in formula (18) we obtain:
a 77/7 2 V3, ; (23)
: h
Where At h is assumed to be - R/' (24)
3.3. The final accuracy of height measurements
Por accuracy of the identification we finally obtain
i^ (25)
and
Shi „AM h' (26)
Total standard error in height measurements is obtained by utlizing formulas (18), (23) and (26):
(s, )* j?
hp = Bar) + (Spa) + (5g) (27)
Average accuracy within the normal stereo-model is obtained by double integration; on the x-axis between
the limits O and b, and on the y-axis between the limits O and d.
x-b yzd
S 2 S 2
beds T - — | ax dy (28)
x=0 y=0 ^
2 2 2
b à 4Th
=D qp tT Qwo* (29)
80 "Up *1uu oy
8
The variation of n for different values of e is demonstrated in fig 4. For e - 0,5 d & minimum
#H
value of 2.00 is obtained. The same value is also obtained for e = - 0.24 d.
In comparison it can be said that in conventional stereo-models (relative-orientated in points 11, 15,
31, 35, 51 and 55 and with elevation control-points in 11, 13, 15, 51, 53 and 55) the value 1.93 is
obtained for the corresponding function.
S S
w
For e = 0.5 d minimum values of = Ix| and £3 have been obtained.
rl
S
>
The appearance of function has been designed in fig 5 for the value e = 0,5 d. About the same
®
accuracies are obtained for e = - 0.25 d.
For comparison we have in fig 6 designed the appearance of the function for e - - d and also in fig 7
designed the appearance of the corresponding function for e - - d and three of the elevation control-
points in 51, 53 and 55 instead of 31, 33 and 35 (the conventional stereo-model). Cf. [5] .
literature:
1) Hellert, B., Stockholm 1956: Uber die Genauigkeit der Luftphotogrammetrie.
2) Kvarnbrink, T. - Thunberg A., Stockholm 1956: A method for the Investigation of Ground Visibility
in Aerial Photography over Forest Areas.
3! Ühlin, K.A., Stockholm 1958: Noggrannheten vid extrapolerad fotogrammetrisk hójdm&átning (The Accuracy
of Extrapolated Photogrammetric Height Measurements).
