System Calibration Procedure
A selection was.made of 39 gridpoint inter-
sections in the gridplate replica, which were
as symmetrically placed as was allowed by the
arrangements of gaugeblocks on that replica
at that time. This arrangement is shown in
Fig 6, being a contact-scale stereogram of
the calibration body. Observations were recor-
ded on each of the selected 39 gridpoints and
on each of the visible surfaces of the gauge-
blocks.
The observations on the gaugeblocks were used
simply to determine a relationship between the
planimetric and the height scales of the af-
fine model. The gridplate observations were
"adjusted to control" taking the planimetric
coordinates of the selected points to vary
both in X and Y between 200 and 400 mm and
attributing a Z value throughout of 200 mm.
The observations were recorded automatically
in the PDP 11/34 which is online to the WILD
A.10 Autograph. These observations were fed
into the Department's Cyber 172 computer,
along with the control values, using a stan-
dard photogrammetric adjustment programme to
compute the standard deviations of the obser-
vation series and to list the residuals.
The results of this calibration study are
tabulated in Table 1. It may be seen that the
difference between the planimetric and the
height scales is quite small, suggesting that
the principal distance adopted for the resti-
tution of the model was close to that of the
image acquisition system, meaning that the
affine deformation was almost negligible.
The standard deviations in X and Y, respec-
tively of 0.11 and 0.15 mm in the scale of
the obejct photographed, seem quite appropri-
ate. The standard deviation in Z, of 0.43 mm
might suggest that the gridplate replica, which
was produced on a dimensionally stable film
but was not held down by such a device as a
vacuum frame,may not have been very flat.
However, The objective was to develop a method
and procedure which may warrant refinement,
not to come up with an ultimate answer at the
first attempt.
TABLE 1
ORIENTATION ELEMENTS:
INSTRUMENTAL SETTINGS ON WILD A.10 AUTOGRAPH,
FOR OBSERVATION OF CALIBRATION BODY NEGATIVES
2 & 3.
PRINCIPAL DISTANCE: 160 mm. BASE: 30 mm
MODEL SCALES: X,Yys 1:2.23. 2: 1:2.33.
Z conversion factor: 0.958.
Z conversion multiplier:
0.43, to read mm.
Kappa. Phi. Omega.
L.H. 99.88 97.490 100.00
R.H. 99.91 102.329 100.00
p. = 30.00 b - 0.00 p - 190.09
x y Z
(The Z range for the observations was about
400 mm.)
DATA REGARDING CALIBRATION BODY AND NEGATIVES:
Gridplate dimensions in replica: 200 x 200 mm
Replica dimensions in negative: 36.65 x 36.70mm
Approximate negative scale is thus: 1:5.45.
(It is to be noted that the stereogram used
for the observation of the calibration body
comprised one "central" Photograph and one
photograph tilted through about four degrees.
The convergence measured by the difference
between the phi settings of the left-hand and
right-hand projectors of the A.10 is 4.839
grades.)
106