(D
‘®
ra”
ween
athe-
It
an
cen
heidt
ver-
the
that
1k of
the
para-
any
ly.
Photogrammetric refraction and earth curvature have been allowed for
throughout the whole experiment in a standard mode.
52 CHOICE OF MATHEMATICAL MODELS
Calibrations have been computed along the lines described in a
previous paper (Kupfer 1976 a). As a standard procedure four strips of
three photographs each at 1:11,000 scale as described before underwent
bundle adjustment, five groups of points in the above mentioned array
serving as full control, the remaining points as check.
(1) Inner orientation was taken from laboratory calibration of the NRC
as communicated to us at March 24, 1975:
{ec}! Calibrated focal length only (this is indicated by "ec" in à
tables); four réseau crosses at the centre of the image borders
served for definition of image coordinate system;
{+} Calibrated focal length as well as radial and tangential distor-
tion (coordinate centre as before).
(2) Réseau information:
{-} Calibration without réseau information;
{+} Calibration with réseau information. Réseau crosses assumed to
be in ideal position (some remarks on deviations are found in
(Fllenbeck 1976)). Image deformation according to rêseau infor-
mation removed by least squares interpolation including four to
six réseau crosses surrounding a point group.
(3) Polynomials for parameterization of partial field calibration.
Parameters chosen give deviations of image coordinates from ideal
position after central projection:
{0} Computation without any additional parameters. Central projec-
tion only.
{1} "Optical" polynomials
2 2 =
dx = x{a {r -r y *tgí(r-r) *....-«b,(r^ t2x^) + b, 2xy
3 O 5 o 1 2
2 2
+ > we 4 + +2
: ac(r T) b,2xy bat? y)
dy = y(a4(r - Tr
2
where r^ 95 x *
2
and r, with the constraint dr ^ = dx” + dy = 0 to be chosen
et will, e.g. y um 90 mm.
TRITT. P
Moss] d. a rud
d
DUAE mad
nun
—
Rl
ia
uisa.
Lus
EEE
Te
TEESE
A0
n°4
E
4
Ce,
ne
©
A
^ri
x
E
ij
^
wi
va
ti
DR
i
E
