Image Processing Reference
In-Depth Information
Based on the fuzzy set, the membership degree of the
IF
image is com-
puted from Chaira's
IF
generator as
−
+−
1
μ
()
g
(5.15)
IFS
A
μ
()
g
=−
1
A
λ
1
(
e
1
) ()
μ
g
A
with λ
> 0.
Using Chaira's fuzzy negation, φ(
x
) = (1 −
x
)/1 + (
e
λ
− 1)
x
, λ
> 0, the non-
membership degree of the
IF
image is computed as
=
(
)
IFS
IFS
ν
(;)
g
λ ϕμ λ
( ;)
g
A
A
or
IFS
−
+−
1
μ
()
g
(5.16)
IFS
ν
()
g
=
A
λ
+
1
IFS
1
(
e
1
) ()
μ
g
λ in Equation 5.16 in the denominator is changed to λ
+ 1, implying that
1
1
+−
λ
IFS
IFS
λ
μ
. With the change in λ,
the non-membership degree,
ν
IFS
(
, will change but will still follow
the condition
ν
(
e
1
) ()
μ
g
is changed to
1
1
+−
(
e
+
1
) ()
g
A
A
IFS
()
1
IFS
()
g
≤−
μ
g
. This is done to obtain a better contrast-
A
A
enhanced image. So,
⎛
⎞
11
1
1
−−
−
+−
μ
()
g
A
⎜
⎟
λ
(
e
1
) ()
μ
μ
g
⎝
⎠
A
IFS
ν
(;)
g
λ
=
A
⎛
⎞
1
−
(
g
)
λ
+
1
A
1
+−−
(
e
1 1
)
⎜
⎟
1
(
λ
1
) ()
+−
e
μ
g
⎝
⎠
A
1
()
−
+− +−
μ
g
(5.17)
A
=
1
(
λ
1
) () (
λ
+
1
1
) ()
λ
e
μ
g
e
μ
g e
A
A
1
()
−
μ
g
A
=
1
+
(
e
21
λ
+
−
1
) ()
μ
g
A
The hesitation degree is computed as
(5.18)
IFS
IFS
IFS
π
(;)
g
λ
=−
1
μ
( ;) (;)
g
λ
−
ν
g
λ
A
A
A
To obtain the optimum value of λ, the
IF
entropy is used.
The
IF
entropy is calculated as
N
N
1
∑
∑
()
()
1
−
π
g
IE A
()
=
π
ge
Aij
⋅
IFSij
N
j
=
1
i
=
1
Search WWH ::
Custom Search