-mail:
scientists.
e sensing
site. Also
ata better
tral data.
generate
bles very
'ration of
] spectral
n and IR
simulated
he Green
he actual
rable.
nan eye.
n virtual
only in
ereas the
QUICK
, red and
the blue
age. In
spectral
ral color
he same
e easily
not have
er to get
simulate
to get a
lone by
rng the
the full
f targets
of IRS-
LISS-III , Quick Bird along with Hyderabad data of IRS-1A
LISS-III has been used. The details areas follow:
Area Satellite Sensor P/R DOP
Hyderaba | IRS 1A LISS II 25/56 27.03.
d 2001
Bangalore | QUICK- MS SAMPL
Bird E
2112:
Chennai IRS-1D LISS-IH | 102/65 98
Methodology:
In satellite remote sensing to display the color composite
: images, three primary colors (red, green and blue) are used.
When these colours are combined in various proportions, they
produce different colors in the image. In satellite imagery the
colors depend on the bands chosen for making the color
composite. Normally in satellite imagery the concept of false
color is common. In false color composite the vegetation is
always in the shades of red. Whenever such images are shown
to the common man or to the people who are new to the remote
sensing data the basic question asked is “ why the vegetation is
red in color”. In order to bring remote sensing to the common
man it is required to show the data in the form they understand
It.
FALSE CO LOUR COMPOSITE
In satellite remote sensing the visual interpretation is carried
out using False color composite, commonly known as FCC
image. NIR, Red and Green spectral bands are assigned the R,
G, B colors to generate FCC. In FCC the color of the target
does not resemble to its actual color but the different spectral
signatures are differentiable. (Ref: Producing remotely sensed
images — Sea space Remote sensing tutorial — 5)
A very common composite scheme for displaying the LISS
multi spectral image is shown below
R = band 4 (NIR)
G = band 3 (red)
B = band 2 (green)
NATURAL COLOUR COMPOSITE
Natural color image in the other words TCC (True Color
Composite) we need to retain the original color of the targets
hence if blue band is available then Blue will be assigned to B,
Green will be assigned to G and red will be assigned to R.
Satellites like IRS-1A/1B, Landsat —5 TM have the sensors in
blue band so image can be assigned respectively to R, G and B
colors for display. In this way the colors of the resulting color
composite image resembles closely what the human eyes
would observe. Satellites like IRS-1C, 1D sensors do not have
a blue band. The four bands correspond to Green (0.52 p —
0.59 p) Red (0.62 p — 0.68 p), Infra red (0.79 p — 0.89 ui) and
SWIR (1.55u- 1.7p). Therefore to make a natural color
composite we need to simulate the blue band.
IAPRS & SIS, Vol.34, Part 7, “Resource and Environmental Monitoring", Hyderabad, India,2002
Natural color
(1.2.3)
IN False color
(4,3,2)
Figure I: IRS 1A LISS II Band Combinations
Simulation of the blue band
The IRS 1A LISS II sensor has four bands Blue ( 0.45 —
0.52) , Green ( 0.52 — 0.60) , Red ( 0.63 — 0.69) , Infra Red (
0.76 — 0.9) (Figure 1.) microns respectively. The Green, Red
and Infra red bands are similar to the Green , Red and Infra red
bands of IRS-1D LISS-III sensor. The blue , green red bands
are in the visible range of the electromagnetic spectrum and are
highly correlated.( Basics of Remote sensing by Lilisand , T.M
Keifer) Therefore if a correlation between the blue , green and
red band is worked out the same can be used to IRS-LISS III
data to simulate the blue band using the green and red bands.
The following steps have been carried out to get the
correlation between the bands.
The IRS-1D and IRS-1A data of Hyderabad area has been
selected of close by dates. The IRS-1A and IRS-1D data sets
were registered. DN values of different targets are noted for
blue and green bands of LISS II sensor and they are converted
into radiance values. Correlation factor was determined for
between these two bands by plotting the values.
Blue Green Blue
(DN (DN (RadianceV Green(Radiance
Value) | Value) alue) Value)
59 24 3.9058 1.6776
75 42 4.965 2.9358
73 38 4.8326 2.6562
86 44 5.6932 3.0756
73 34 4.8326 2.3766
76 37 5.0312 2.5863
82 41 5.4284 2.8659
63 33 4.1706 2.3067
102 51 6.7524 3.9843
71 30 4.7002 2.097
This correlation factor is not uniform for all targets. It varies
with the type of the targets. Because of this radiance values
were classified as three categories of the dynamic range as
Low, Medium and maximum radiance and then correlation
factor was determined separately for each class. The green
band was multiplied by these factors by using the spectral math
routine of the image analysis software. Then to get the exact
