Image Processing Reference
In-Depth Information
b
127
Lab
L ΄ a ΄ b ΄
α
a
-128
127
Constant L
-128
FIGURE 7.61
Gamut mapping for constant lightness and hue.
q
( L
2
r ¼
50
)
þ a 2
þ b 2
(
7
:
100
)
a ¼ tan 1 b
a
L
50
a 2
u ¼ tan 1
and
p
þ b 2
In these equations, r is the distance from the gamut centroid to the given
color point,
a
is the hue angle with the dynamic range from 0
8
to 360
8
and
u
.
2. Find the intersection of the gamut with a straight line connecting the input
color point having coordinates [r 0 a 0 u 0 ] to the centroid of the gamut.
This is achieved by searching the gamut boundary points with coordinates
that fall into the range
is the angle in the constant
a
plane from 0
8
to 180
8
. The average of these boundary
points is denoted by [r ave a ave u ave ]. Now the desired color point is inside
the gamut if r 0 < r ave , otherwise it is outside the printer gamut.
a 0 Da
and
u 0 Du
Once we determine that the color is outside the printer gamut, we map it to the
nearest point on the gamut boundary that best satisfies the conditions shown in
Equation 7.99.
7.6.5 M ERIT -B ASED G AMUT M APPING
There is no unique gamut-mapping method that can satisfy requirements like pleasing
color, contrast, lightness, chroma, hue, etc. More than 90 gamut-mapping algorithms
[111] can be used for mapping colors to a suitable region of the printer gamut. Some
offer feature enhancements in one region of the gamut. Others do the same in different
regions of the gamut. As such, device designers generally compromise in the gamut-
mapping functions they wish to employ in their respective color management systems.
Merit-based feedback system used for gamut mapping offers a way to automatically
select the gamut-mapping algorithms from the library of algorithms to optimize a
merit function [134]. In this method, all node colors are clustered in different regions
of interest within the color space and each cluster is associated with at least one
 
Search WWH ::




Custom Search