Method for enhancing local contrast of an image
A method for enhancing local contrast of an image. The method comprises receiving a component of an image, computing a wavelets decomposition of the received component into wavelet coefficients, determining at least three subsets of wavelets coefficients based on a first threshold, a second threshold and computed wavelet coefficients, modifying the determined at least three subsets by applying at least a sharpening transformation on wavelet coefficients of at least one subset in said at least three subsets, recombining the at least three subsets into final wavelet coefficients, and reconstructing a new component based on the final wavelet coefficients.
1. A method for enhancing local contrast of an image, wherein the method comprises:
receiving a component of the image;
computing wavelet coefficients based on a wavelet decomposition of the received component;
computing a first threshold and a second threshold based on a brightness of the image;
determining at least three subsets of wavelet coefficients based on the first threshold, the second threshold and the computed wavelet coefficients;
modifying the at least three subsets via applying a sharpening transformation on at least one of the at least three subsets;
recombining the at least three subsets into final wavelet coefficients; and
reconstructing a new component based on the final wavelet coefficients.
2. A method according to claim 1 , wherein determining the at least three subsets of wavelet coefficients is a function of the absolute value of the computed wavelet coefficients.
3. A method according to claim 1 , wherein the received component is a luminance component.
4. A method according to claim 1 , wherein the first threshold is a function of a noise threshold of the image representing the amount of noise in the image.
5. A method according to claim 1 , wherein the second threshold is a function of a boundary between medium contrast and high contrast differences.
6. A method according to claim 1 , wherein the computed wavelet coefficients are d j k [m,n] and a D [m,n], j being an integer ranging from 1 to an integer D, k being an orientation index, and n and m being coordinates in the image, wherein the first threshold is T n , wherein one subset of the at least three subsets comprise wavelet coefficients low j k [m,n],
wherein
low
j
k
[
m
,
n
]
=
{
d
j
k
[
m
,
n
]
if
d
j
k
[
m
,
n
]
<
D
1
(
T
n
)
[
m
,
n
]
0
otherwise
,
and wherein D 1 (.) is a distortion function.
7. A method according to claim 1 , wherein the computed wavelet coefficients are d j k [m,n] and a D [m,n], j being an integer ranging from 1 to an integer D, k being an orientation index, and n and m being coordinates in the image, wherein the first threshold is T n , wherein the second threshold is T h , wherein one subset of the at least three subsets comprise wavelet coefficients high j k [m,n], wherein
high
j
k
[
m
,
n
]
=
{
0
if
c
j
k
[
m
,
n
]
≤
D
2
(
T
h
)
[
m
,
n
]
λ
(
c
j
k
[
m
,
n
]
-
D
2
(
T
h
)
[
m
,
n
]
)
if
D
2
(
T
h
)
[
m
,
n
]
<
c
j
k
[
m
,
n
]
≤
λ
λ
-
1
*
D
2
(
T
h
)
[
m
,
n
]
c
j
k
[
m
,
n
]
if
c
j
k
[
m
,
n
]
>
λ
λ
-
1
*
D
2
(
T
h
)
[
m
,
n
]
wherein D 2 (.) is a distortion function,
wherein λ is a real number, and wherein c j k [m,n] is a coefficient between 0 and d j k [m,n].
8. A method according to claim 1 , wherein the computed wavelet coefficients are d j k [m,n] and a D [m,n], j being an integer ranging from 1 to an integer D, k being an orientation index, and n and m being coordinates in the image, wherein the first threshold is T n , wherein one subset of the at least three subsets comprise wavelet coefficients low j k [m,n],
wherein
low
j
k
[
m
,
n
]
=
{
d
j
k
[
m
,
n
]
if
d
j
k
[
m
,
n
]
<
D
1
(
T
n
)
[
m
,
n
]
0
otherwise
,
wherein D 1 (.) is a distortion function,
wherein the second threshold is T h ,
wherein one subset of the at least three subsets comprise wavelet coefficients high j k [m,n],
wherein
high
j
k
[
m
,
n
]
=
{
0
if
c
j
k
[
m
,
n
]
≤
D
2
(
T
h
)
[
m
,
n
]
λ
(
c
j
k
[
m
,
n
]
-
D
2
(
T
h
)
[
m
,
n
]
)
if
D
2
(
T
h
)
[
m
,
n
]
<
c
j
k
[
m
,
n
]
≤
λ
λ
-
1
*
D
2
(
T
h
)
[
m
,
n
]
c
j
k
[
m
,
n
]
if
c
j
k
[
m
,
n
]
>
λ
λ
-
1
*
D
2
(
T
h
)
[
m
,
n
]
wherein D 2 (.) is a distortion function,
wherein λ is a real number,
and wherein c j k [m,n] is a coefficient between 0 d j k [m,n].
9. A method according to claim 1 , wherein the computed wavelet coefficients are d j k [m,n] and a D [m,n], j being an integer ranging from 1 to an integer D, k being an orientation index, and n and m being coordinates in the image, wherein the at least three subsets comprise wavelet coefficients j k [m,n], med j k [m,n], and high j k [m,n], and wherein d j k [m,n]=low j k [m,n]+med j k [m,n]+high j k [m,n].
10. A method according to claim 1 , wherein recombining the at least three subsets further comprises adding wavelet coefficients of the at least three subsets for the same coordinates in the image.
11. A device for enhancing local contrast of an image, wherein the device comprises:
an interface configured to receive a component of the image;
a circuit configured to compute wavelet coefficients based on a wavelet decomposition of the received component;
a circuit configured to compute a first threshold and a second threshold based on a brightness of the image;
a circuit configured to determine at least three subsets of wavelet coefficients based on the first threshold, the second threshold and the computed wavelet coefficients;
a circuit configured to modify the at least three subsets via applying a sharpening transformation on at least one of the at least three subsets;
a circuit configured to recombine the at least three subsets into final wavelet coefficients; and
a circuit configured to reconstruct a new component based on the final wavelet coefficients.
12. A non-transitory computer readable storage medium having stored thereon instructions that, when executed, cause a processor of a device to:
receive a component of an image;
compute wavelet coefficients based on a wavelet decomposition of the received component;
compute a first threshold and a second threshold based on a brightness of the image;
determine at least three subsets of wavelet coefficients based on the first threshold, the second threshold and the computed wavelet coefficients;
modify the at least three subsets via applying a sharpening transformation on at least one of the at least three subsets;
recombine the at least three subsets into final wavelet coefficients; and
reconstruct a new component based on the final wavelet coefficients.