<s>
Multispectral	B-Algorithm
remote	I-Algorithm
sensing	I-Algorithm
is	O
the	O
collection	O
and	O
analysis	O
of	O
reflected	O
,	O
emitted	O
,	O
or	O
back-scattered	O
energy	O
from	O
an	O
object	O
or	O
an	O
area	O
of	O
interest	O
in	O
multiple	O
bands	O
of	O
regions	O
of	O
the	O
electromagnetic	O
spectrum	O
(	O
Jensen	O
,	O
2005	O
)	O
.	O
</s>
<s>
Subcategories	O
of	O
multispectral	B-Algorithm
remote	I-Algorithm
sensing	I-Algorithm
include	O
hyperspectral	O
,	O
in	O
which	O
hundreds	O
of	O
bands	O
are	O
collected	O
and	O
analyzed	O
,	O
and	O
ultraspectral	O
remote	O
sensing	O
where	O
many	O
hundreds	O
of	O
bands	O
are	O
used	O
(	O
Logicon	O
,	O
1997	O
)	O
.	O
</s>
<s>
The	O
Iterative	O
Self-Organizing	O
Data	O
Analysis	O
Technique	O
(	O
ISODATA	O
)	O
algorithm	O
used	O
for	O
Multispectral	B-Algorithm
pattern	I-Algorithm
recognition	I-Algorithm
was	O
developed	O
by	O
Geoffrey	O
H	O
.	O
Ball	O
and	O
David	O
J	O
.	O
</s>
<s>
There	O
are	O
several	O
multispectral	B-Algorithm
remote	I-Algorithm
sensing	I-Algorithm
systems	O
that	O
can	O
be	O
categorized	O
in	O
the	O
following	O
way	O
:	O
</s>
<s>
Multivariate	O
statistical	O
parameters	O
(	O
means	O
,	O
standard	B-General_Concept
deviations	I-General_Concept
,	O
covariance	O
matrices	O
,	O
correlation	O
matrices	O
,	O
etc	O
.	O
)	O
</s>
<s>
All	O
pixels	B-Algorithm
inside	O
and	O
outside	O
of	O
the	O
training	O
sites	O
are	O
evaluated	O
and	O
allocated	O
to	O
the	O
class	O
with	O
the	O
more	O
similar	O
characteristics	O
.	O
</s>
<s>
This	O
vector	O
would	O
contain	O
the	O
brightness	O
values	O
for	O
each	O
pixel	B-Algorithm
in	O
each	O
band	O
in	O
each	O
training	O
class	O
.	O
</s>
<s>
The	O
mean	O
,	O
standard	B-General_Concept
deviation	I-General_Concept
,	O
variance-covariance	O
matrix	O
,	O
and	O
correlation	O
matrix	O
are	O
calculated	O
from	O
the	O
measurement	O
vectors	O
.	O
</s>
<s>
This	O
process	O
consists	O
in	O
a	O
series	O
of	O
numerical	O
operations	O
to	O
search	O
for	O
the	O
spectral	O
properties	O
of	O
pixels	B-Algorithm
.	O
</s>
<s>
In	O
the	O
second	O
pass	O
,	O
a	O
minimum	O
distance	O
to	O
means	O
classification	O
algorithm	O
is	O
applied	O
to	O
the	O
dataset	O
,	O
pixel	B-Algorithm
by	O
pixel	B-Algorithm
.	O
</s>
<s>
Then	O
,	O
each	O
pixel	B-Algorithm
is	O
assigned	O
to	O
one	O
of	O
the	O
mean	O
vectors	O
created	O
in	O
the	O
first	O
step	O
.....	O
</s>
<s>
The	O
ISODATA	O
algorithm	O
is	O
a	O
modification	O
of	O
the	O
k-means	B-Algorithm
clustering	O
algorithm(overcomes the disadvantages of k-means )	O
.	O
</s>
