centre and can then be considered as just
one more point in the main photo scheme, If
the relationship is not known, then it can of
course be readily established if there are
enough common points, The angle of parallax
Subtended by the two camera centres at the
distant ground point will normally be completely
negligible.
If the heavenly body has such a low elevation
that it is not convenient to take the intercept
with the datum plane, then any convenient Xzconst
or Y=const plane may be used; in this case, it
probably will also be inconvenient to take the
intercept with the photo plane of the main camera,
This need cause no difficulty, but the full formula
must be used when computing the constant term for
the observation equations,
(e)
Compass or other Heading. It will be assumed
that the bearing has been converted to its
value in the ground coordinate system, In
addition to the principal point, take a second
"dummy" photo point at the end of the x-axis;
there will now be (2p+1) observation equations
for a photo with p normal photo points, If now
the intercept with the datum plane of the ray
through this second dummy point is (95:9): then
the heading of the photo has tangent (T-R)/(S-Q),
and one has the usual choice of including an
additional observation equation to express the fact
that this is the, given, tangent of the known
angle, or of eliminating one of the variables
to make the condition rigid,
(f) Rounds of Theodolite Angles. These could be
observed at a ground photo point to other photo points,
or known triangulation points, they could include
sun or star shots, or rays to air stations - the
point occupied need not even be a photo point at
all, In all cases, the most satisfactory procedure
would be to have both horizontal and vertical angles
observed, and to reconstruct the space (or "sextant")
angles from point to point, After this, a space
angle observed by a theodolite is on exactly the
same footing as a space angle "observed" in an air
photo, and the observation equation coefficients
have exactly the same form in both c ases, It would,
also, of course be possible to make two observation
equations, by keeping horizontal ahd verti cal
measures separate, but problems of weighting would
then arise,
