sion
latching,
hich will
size from
he exact
lering all
listortion
erospace
libration
ribes the
with the
natching
'esented,
tions.
) orthog-
era or a
e angles
tes. For
0 spatial
epresent
owledge.
urement
y of the
value, if
] by the
process
into two
sion and
on. The
e results
ne cases
changes
2 evalu-
. exactly
letermi-
)-arrays,
s. The
etection
intation
MIR co-
out the
MW Radionicter
IKAR-N
MW Panorama
Radiometer
IR Spectrometer
ISTOK- 1 and
VIS/NIR Spectrometer
MOS - ODSOR (inside)
MW Scanning Radiometer
R-40
LIDAR ALISSA (inside)
Priroda Module
Figure 1: PRIRODA module on the MIR station
Microwave (MW) Radiometer
R-30,R-80, R-135
absolute pointing of MOMS-2P. These investigations will
also try to include the star camera on MIR module Quant.
This paper is organized in two main sections and a final
conclusion part. The following section describes the lab-
calibration, the equipment, the process and shows the re-
sulting calibration parameters. The next section deals with
the inflight calibration. The principle, the test areas and the
mission scenario are outlined. A technique for the automated
identification of a large number of ground control points,
via an approach for multi-scale matching using line imagery
and existing DOI is described. Subsequently, the results of a
simulation study are presented, showing to which extend the
camera parameters can be determined by photogrammetric
means under realistic conditions.
2 LAB-CALIBRATION
Extensive alignments and measurements have been performed
in the laboratory to characterize the MOMS-2P optics mod-
ule. The main tasks were:
e Alignment of the CCD detectors into the optimum fo-
cal plane of each lens assembly. Since the lenses are
nearly diffraction limited (i.e. the physical limit), the
tolerance for focus adjustment was « 20 um.
e Lateral adjustment to obtain exact superposition of all
detector arrays. The ground footprints of all CCD-
arrays (i.e. the geometrical projection of the CCD's
through the lenses onto the earth surface) shall be cor-
related with sub-pixel accuracy (« 0.1 pixel) which cor-
responds to alignment and stability tolerances of 1 um
at detector level.
e Adjustment of the projected image size between all
channels, which implies that the focal length of all five
lenses must be matched with sub-pixel accuracy. Note
that the relative focal length accuracy is 3 - 107°.
e Measurements of detector geometry and the rela-
tion of the individual channels together, again with
Model cf the Priroda Module on the MIR Station
91
Figure 2: Optics module of MOMS-2P
sub-pixel accuracy.
e Relative and absolute radiometric calibration.
An important aspect for all on-ground testing activities was
the fact that the instrument shall be characterized for vac-
uum conditions. The optics module, however, exhibits signif-
icantly different performances when operated in vacuum or
in air. This effect is due to the small but finite refractive
index difference between air (ne = 1.000297) and vacuum
(n. — 1.000000). A natural approach to cope with this prob-
lem would be to place the entire optics module and all asso-
ciated test equipment in a vacuum chamber. However, the
drawbacks would be very inconvenient such that there is no
direct access to the equipment and to adjustment tools other
than by remote (computer) control, that the required facilities
are very expensive and that the resulting logistic difficulties
would stretch the time scale extremely.
The dedicated solution for MOMS-2P was found in air spaced
doublets which are used only for ground testing. For each
lens assembly an individual doublet was designed and manu-
factured. Attached to the lens assembly, the doublets exactly
compensate the change of focal length and of focus posi-
tion when going from air to vacuum. This approach was the
prerequisite to predict and characterize the overall system
performance for vacuum conditions.
2.1 Calibration Procedure
The calibration procedure was based on the philosophy to
design the test setup for maximum commonality, i.e. to re-
duce the number of dedicated equipment and to reduce the
number of required changes of the test setup for individual
measurements. Basically, the test setup for MOMS-2P has
been dictated by the very high geometric accuracy require-
ments. For this reason, all channels (and therefore all lens as-
semblies) were stimulated simultaneously by the same source.
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B1. Vienna 1996
