- •4 Color appearance models
- •4.1 General requirements for color appearance models
- •4.2 Cielab Model
- •Inverse conversion equations
- •4.3 Cieluv Model
- •Inverse conversion equations
- •4.4 Ciecam02 Model
- •Categorical viewing conditions settings for the model
- •Unique hue data for the calculation of hue quadrature
- •Inverse conversion equations
- •4.5 Modification of ciecam02 by Luo et al.
- •Table 4.1
- •Values of lightness for spectral colours
- •Table 4.2
- •Values of colorfulness for spectral colours
- •Table 4.2 (End)
- •Table 4.3 Values of criterion
- •Table 4.4 (End)
- •Таble 4.7 Values of distance between position of points of monochromatic colors of chromaticity diagram for combined adapting luminance and surround for stimulus luminance, equal 10 cd/m2
- •Table 4.9 Values of distance between position of points of monochromatic colors of chromaticity diagram for stimulus luminance, equal 20 и 200 cd/m2 for stimulus luminance, equal 50 cd/m2
- •Table 4.11 Values of distance between position of points of monochromatic colors of chromaticity diagram for combined adapting luminance and surround for stimulus luminance, equal 50 cd/m2
- •Table 4.12
- •Figure 4.19 ‒ Occurrence of image impairment in dependence of colour deflection levels
- •4.6 High-Luminance Color Appearance Model
- •Inverse model computation:
Unique hue data for the calculation of hue quadrature
|
Red |
Yellow |
Green |
Blue |
Red |
|
1 |
2 |
3 |
4 |
5 |
|
20.14 |
90.00 |
164.25 |
237.53 |
380.14 |
|
0.8 |
0.7 |
1.0 |
1.2 |
0.8 |
|
0.0 |
100.0 |
200.0 |
300.0 |
400.0 |
The achromatic response is:
(4.48)
The lightness is:
(4.49)
The brightness is:
(4.50)
The chroma is:
(4.51)
here
(4.52)
The colorfulness is:
(4.53)
The saturation is:
. (4.54)
CIECAM02
includes three attributes in relation to the chromatic content:
chroma (C),
colorfulness (M)
and saturation (s).
These attributes together with lightness (J)
and hue angle (h)
can form three colour spaces
,
where
|
|
|
|
|
|
CIECAM02
chromaticity diagram
CIECAM02
chromaticity
diagram is presented in Figures 4.1, 4.2 and 4.3 [4.6]. The figures
demonstrate the dependence of colour appearance on adaptation level
LA
and relative luminance level Y.
It is seen from this figures that a variation of adapting luminance
level may lead to significant colorfulness variation, and that
variation of colorfulness increases with stimulus relative luminance
growth.
Inverse conversion equations
First
it’s necessary to compute surround-dependent parameters and
achromatic response for reference white
by using forward conversion. Lightness value
can be obtained from brightness by using the following equation:
(4.55)
The chroma is:
(4.56)
If starting from lightness, the brightness can be calculated:
(4.57)
If saturation is known the chroma is defined as follows:
(4.58)
If hue quadrature data is available the hue angle can be calculated using unique hue data for the calculation of hue quadrature as:
(4.59)
here
,
if
,
otherwise
(4.60)
The eccentricity factor is:
(4.61)
The achromatic response is:
(4.62)
The
following relationships are used for obtaining
and
values:
(4.63)
(4.64)
(4.65)
(4.66)
if
,
then:
(4.67)
(4.68)
(4.69)
if
,
then:
(4.70)
(4.71)
(4.72)
The cone responses are:
(4.73)
(4.74)
If
any of the values of
are negative, then the corresponding value
must be negative.
(4.75)
(4.76)
(4.77)
Tristimulus values of the sample then are:
(4.78)
4.5 Modification of ciecam02 by Luo et al.
All
the colorimetric assessments based on CIECAM02 are usually expressed
in
or
spaces. As it was shown [4.7 et al.], usage of
‒space
gives more accurate predictions of color appearance. The following
modifications of this space for large (CAM02-LCD), small (CAM02-SCD)
and both small and large (CAM02-UCS) color differences (see Section
6) were proposed:
(4.79)
(4.80)
(4.81)
The coefficients for each version of UCS based upon CIECAM02 are the following:
-
Version of space
CAM02-LCD
CAM02-SCD
CAM02-UCS
0,77
1,24
1,00
0,007
0,007
0,007
0,0053
0,0363
0,0228
Correlation
of values
,
is presented on Figures 4.4 and 4.5.
As follows from published results of studies [4.7 et al.], the estimations got with the use of these modifications show the best correlation with all available data on colour appearance and can be considered as basis for the further studies directed to progress of the television and related video applications and to progress colour appearance models for their use as part of the systems of image quality evaluation, in particular, evaluation of colorimetric quality.
It should be noted necessity of further colorimetric studies, related to the television specific taking into account characteristics of capture and reproduction, influencing on image quality, conditions of scenes capture and reproduction etc.
The results of testing published have shown that predictions obtained by using CIECAM02-based color spaces best match all available color appearance data and can be considered to become a base for further research work on development of TV and related video systems and on development of color appearance models for implementation them as the part of image quality assessment systems, particularly colorimetric quality assessment.
It should be noted, that there’s a need in further colorimetric research according to television specificity with taking into account the specific features of TV image capture and reproduction, influencing quality, the conditions of scene capture and reproduction and so on.
CAM02-UCS
(
)
chromaticity diagram
CAM02-UCS ( ) chromaticity diagram [4.6] is presented in Figures 4.6, 4.7 and 4.8. The figures demonstrate the dependence of colour appearance on adaptation level LA and relative luminance level Y in Luo et al. colour space.
In
Table 4.1 the values of lightness
of spectral color samples with given reference luminance
for different adapting luminance
levels are presented. The Table shows in what degree perceived
lightness
depends on sample colours. It also shows that the dependence of
on
is insignificant.
