(g) Spirit Level or other Indication of the
Vertical,
This may be a device in the air camera, at the
air station, or the implicit observation
made by levelling the theodolite at the ground
point before taking a round of angles, but in either
case the procedure is the same. One takes the
plumb point, with its projection (U,V,0) on the
datum plane, # as an additional "photo point",
leading to two additional observation equations;
in the case of the air camera, the pair of "photo
coordinates" produced for this point will depend
on the nature of the device in the air cameras
in the case of the ground theodolite, the two
"space angles" in question will, of course, be
any convenient two of the vertical angles
observed at the point (corrected for refraction,
since this is so much more apparent for vertical
angles observed from ground point to ground point).
It may be noted that, if deviation of the vertical
were known, this could also be allowed for at
this point, by varying the coordinates of the
projection of the plumb point, and either holding
it rigidly in relation to (U,V,-D)s or making it
a flexible control to obtain a best mean fit,
leading perhaps to a better value of the deviation.
Note: This may perhaps be a convenient
place to point out that this method
seems in general applicable to ordinary
ground triangulation schemes, quite
apart from photogrammetry
(h) Air station fixed by various methods from the
ground. This is clearly identical with (a) (Plan
and Height Control), since one merely equates the
air station coordinates, rigidly or flexibly,
to the known values.
+ The datum would be too close, in the ground case,
hence any convenient plane Z- -D is then chosen,
6, Inverting Large Band Matrices
( À LL
the limit of the capacity of the small drum Pegasus
ronic computer to invert the matrix, by Gauss elimination,
passing data only once through store, consideration of an
alternative method is given in Appendix D.
Since, when working with about 80 variables, one is just
+
11
t!
— —— ———
