Farnborough, U.K., gave an estimated spatial reso
lution at 4.5 m in range and 4.0 m in azimuth. For
the data correlated at the Deutsche Forschungs und
Versuchsanstalt fur Luft und Raumfahrt (DFVLR),
Oberpfaffenhofen, West Germany, the spatial resolu
tion was in the order of 5 m or higher (Smith, op.
cit.).
Spatial resolution for the original optical pro
cessed imagery at the Environmental Research Insti
tute (ERIM), Michigan, USA, was specified as being
in the order of 1.5 m in range and 2.1 m in azimuth
direction (Intertech 1981), calculated on the basis
of the properties of the imaging system.
EXPERIMENT WITH OIL DRUMS AS SPATIAL RESOLUTION
TARGETS
In many radar surveys, where images are optically
correlated, it has been found difficult to check the
actual ground resolution of the system with respect
to the resolution as specified by the contractor.
The here presented experiment is based on measu
ring the minimum distance between two point targets
with approximately the same response strength which
are still visible as two objects on the image. One
option was to use a number of corner reflectors or
Lunenburger lenses, placed in a grid with variable
grid cell size. These reflectors are often difficult
to procure, however and moreover, they are costly,
especially when a survey is carried out over remote
areas in developing countries. We therefore decided
to test a material easily and cheaply available
everywhere in the world. These consisted of oil
drums, cut in half lengthwise. These half drums
were welded together end-to-end. In this way, con
cave upwards, they formed a perfect cylindrical
reflector of a diameter of 0.7 m and a length of
2.50 m. The corrugation of the drumwalls (amplitude
3 cm) served as resonant elements and strengthened
the return signal from these "drum reflectors" (fig.
1) .
Fig. 1 Drum reflectors-in the test site
The reflectors were oriented with their length
direction either roughly parallel or slightly obli
que to the flight direction. Some of them were in
clined towards the look direction under different
angles to check the influence of orientation on the
strength of the return signal. The spacing between
the drums varied between 1.5 m and 13m. The comple
te configuration is sketched in figure 2. Nine "drum
reflectors" were situated on line (direction 83°)
roughly parallel with the flight direction (73°).
They have a uniform low intensity background clut
ter from a smooth grass surface. Directly south of
this row of drums was a 10 m wide ploughed zone
serving as a fire divide. Three more drum reflec
tors were situated in the grass area between the
ploughed strip and the barbed wire fence along the
road (fig. 1, drums 10 - 12). On the southern side
Fig. 2 Orientation of drum reflectors in the
test site
of the asphalt road, there is another barbed wire
fence, followed by a grass strip a few metres wide
and another 10m wide ploughed fire divide. Farther
south, there is a rather homogeneous grass cover.
The site was flown during a mission on 23 June
1981, With a synthetic aperture radar in X and C
bands. Three overlapping strips cover the test
area. They have been optically processed (high
resolution). No digital data were obtained over
this area.
The northernmost strip 78 is available in X- and
C-bands with depression angles of 32° towards the
site. Strips 77 and 76 are available only in C-band
because of a malfunction of the X-band during the
flight. The depression angles are, respectively,
42° and 70°.
RESULTS OF MEASUREMENTS
On all four radar strips the field of drum reflec
tors could be easily located as bright patches by
the high radar return. The positive film has been
studied under the microscope with 63x enlargement
to differentiate further detail on the transparen
cies.
The smooth asphalt road surface appears as a dark
line (light tone on the negative) as could be ex
pected because of the specular reflection. On both
sides of the road, light toned lines occur repre
senting the barbed wire and/or the shallow depres
sion along both sides of the road. This is followed
by a dark toned line with a smooth surface texture,
caused by the 10 m wide grass strip which separates
the fence from the ploughed fire divide. The fire
divide appears light in tone because of the rough-
Fig. 3 Point scattering of drum reflectors on
strip 77c (optically correlated)