The aerial photographs were taken previously(in 1975)
and were especially useful as the zones being examined have
undergone no change in the years since then, as far as these
groups are concerned.
Available data was supplied in positive black and white
trasparencies in the four Landsat channels ((D)). These were
enlarged electronically at a ratio of 6:1 (produced then about
at 1:166,667) after the data had been expanded radiometrically
so that the centre of the gray scale fell around the density
corresponding to the average relative signature of the agri
cultural species in t he zone at the mostjrigorous phenological
stage prior to ripening |
Radiometric expansion allowed signatures lying in the vicinity
of average signatures of vigorous vegetation to be emphasised
while the lowest signatures (bare ground,urban settlements and
the like)and the higest ones (shores,snow and others) were
"compressed".
During the analysis of each of the available scenes the
combination of channels 4 and 5 were inverted photographically,
so that the non-agricultural coverages in these two channels
had larger signatures than the vegetation ones; by inverting
the process the non-agricultural signatures were "depressed"
and the agricultural ones emphasised. The trasnsparencies of
channels 6 and 7 were then superimposed over the inverted
scene of channels 4 and 5 using a shot of the combination.
As the relations between signatures in channels 6 and 7
are inverted compared to the existing ones in channels 4 and 5
a black and white map emerges which emphasises only the
signatures around which radiometric expansion is centred:
in this case the vigorous vegetation signatures.
the |
6 an
By a
of t
the
one
"dep
is o
obta
the
smal
mead
crop
1976
spec
the
at t
reac
prio
logi
fiel
cere
Scer
char