lir direction, 
: the interband 
; for instance 
the combina- 
lired 9 bands 
r and aft 
:lly regis- 
iplied off-line 
;ize and for 
The module 
i be tilted 
that a compact 
to the total 
'indow in the 
atory air- 
.ESAR camera 
separation for 
minimized by 
void the intro- 
room scan prin- 
he object 
a for sea ob- 
em, the dy- 
ld be deter- 
ance, the 
bserved, the 
of the ima- 
ptics and the 
ty, band width, 
addition the 
role. For land 
the required 
lution of 0.5 
of the mea- 
emperature 
calibration in 
ontrol would 
ents for the 
he same 
ervation 
400-420 nm 
435-455 nm 
510-530 nm 
555-575 nm 
620-640 nm 
675-695 nm 
770-800 nm 
990-1050 nm 
ed twice) 
£ 0,05% for 
all bands 
10 and 20 m 
5.7°. 
Fig. 1 Selection of spectral bands for land and 
sea observation by CAESAR. 
Fig. 2 CAESAR CCD camera module 
The main components are the standard 
objective, the spectral separation, 
bandfilters and CCD detectors and the 
selector electronics. 
From a market survey executed in 1982, three manu 
facturers of CCD's with 1728 detector elements were 
selected on basis of a set of primary criteria. A 
dedicated CCD testbench has been developed by the 
TPD. A detailed test programme has been executed 
with respect to the following characteristics: dark 
current (non-uniformity and temperature dependence), 
detector sensitivity, radiometric linearity and dy 
namic range, detector cross- talk, efficiency of 
signal conversion and charge transport. Finally one 
type of CCD was selected. 
DATA ACQUISITION AND REGISTRATION 
The NLR has developed a family of airborne analog to 
digital data conversion systems for different types 
of sensors like Side-looking Airborne Radar, Infra 
red line Scanner and Video camera. Based on the 
approach of a dedicated analog to digital convertor 
combined with a standardized programmable data for 
matter and an airborne high bit rate recording system, 
Fig. 3 Pushbroom scanning for land observation 
with 3 co-registered (central) channels 
of the modules; for sea observation all 
9 channels (3x3 co-registered) are used. 
for CAESAR the multichannel convertor "CEDIG" has 
been developed. 
The maximum input data rate for the HBR has been 
upgraded to 8.4 Mbits per sec. The CEDIG system is 
programmable with respect to the number of pixels 
per scanline, the number of bits per pixel and the 
integration time. The number of samples per line is 
either 1280 or 1792. Electronic roll correction is 
realized by positioning the 1728 pixels within the 
1792 samples or by selection of 1280 pixels from the 
1728 measurements. 
By changing the integration time the number of scan 
lines per sec with corresponding along-track spatial 
resolutions can be changed stepwise. In such a way 
trade-offs are possible between intensity resolu 
tion, spatial resolution, swath width and number of 
channels. 
REALISATION AND FLIGHT TESTING 
Figure 4 shows the configuration of the CAESAR sensor 
system. Three identical CCD camera modules are moun 
ted on a common base plate and are aligned in paral 
lel in the optical laboratory of the TPD. The three 
modules are protected by means of a box. This down 
looking box is mounted in a support structure. The 
baseplate can be adjusted at the selected tilt 
angle. Beside the three modules, the forward-looking 
module is mounted within the support structure. The 
off-nadir direction can be varied with a number of 
discrete steps. 
The internal temperature of both modules can be 
maintained at a nominal level by means of a flow of 
dry cooled air. This is required in case of opera 
tions in tropical regions. 
The electronics, the inertial navigation platform 
and the high density recorder are mounted in remo 
vable racks inside the aircraft. The first in-flight 
tests have been executed in 1984 for initial perfor 
mance testing of the integrated system and for the 
generation of image data required for the develop 
ment and testing of the dedicated system correction 
(preprocessing) software. Figure 5 presents an 
example of one of the first images. No geometric and 
radiometric corrections have been applied, except 
the real time roll correction. After the first in 
flight tests a series of environmental tests have 
been executed, like vibration tests, temperature and 
electromagnetic interference tests. Some mechanical 
improvements of the support structure and modifica-