 |
Claims  |
|
|
I claim:
1. A method of periodically calibrating a cathode-ray tube display (CRT) to
establish a stored relationship between a set of RGB signal values and
associated CRT intensity values such that perceptually accurate colors are
displayed on the CRT in response to color input signal values, comprising
the steps of:
driving the CRT with an initial RGB signal value that displays a
predetermined color on the CRT;
storing in a memory an initial CRT intensity value associated with the
initial RGB signal value and the corresponding predetermined color
displayed on the CRT;
subjecting the CRT display to a period of use likely to cause a change in
the predetermined color displayed in response to the initial RGB signal
value;
stepping the RGB signal values through a group of values;
storing in a lookup table a sensed current CRT intensity value associated
with each stepped RGB signal value;
finding in the lookup table a current CRT intensity value associated with
the initial RGB signal value;
comparing the current CRT intensity value with the initial CRT intensity
value, and generating a correction factor;
receiving a color input signal value;
altering the color input signal with the correction factor to generate a
corrected RGB signal value; and
driving the CRT with the corrected RGB signal value to display on the CRT a
perceptually accurate color in response to the received color input signal
value.
2. The method of claim 1 in which the stepping step and the storing in a
lookup table step further comprise:
shielding from ambient light a sensed portion of the CRT;
stepping the RGB signal values repeatedly through the group of values; and
storing in the lookup table an average of the sensed CRT intensity values
associated with each repeatedly stepped RGB signal value.
3. The method of claim 1 further including a step of deleting from the
lookup table RGB signal values for which an associated sensed CRT
intensity value is substantially zero.
4. The method of claim 1 further including a step of adding to the lookup
table computed non-zero CRT intensity values corresponding to any RGB
signal values missing from the group of stepped RGB signal values.
5. The method of claim 3 in which the added CRT intensity values are
computed by using a polynomial-based curve fitting calculation.
6. The method of claim 1 in which the storing in a memory step further
includes measuring and storing in an SA matrix a set of CRT color data
representing a relationship between the RGB signal values driving the CRT
and a set of device independent color values,
7. The method of claim 6 further including the steps of:
multiplying the SA matrix by the correction factor to generate a
corrected matrix A relating device independent color values to
associated CRT intensity values;
receiving a particular device independent color value signal;
dividing the particular device independent color value signal by A to
generate a particular CRT intensity value;
finding in the lookup table the particular RGB signal value corresponding
to the particular CRT intensity value; and
driving the CRT with the particular RGB signal value to display on the CRT
a colorimetrically accurate color in response to the particular received
device independent color value signal.
8. The method of claim 7 in which the device independent color values are
CIE tri-stimulus XYZ values.
9. A method of periodically calibrating a cathode-ray tube display (CRT) to
establish a stored relationship between a set of RGB signal values and
associated CRT intensity values such that perceptually accurate colors are
displayed on the CRT in response to CIE tri-stimulus XYZ color input
signal values, comprising the steps of:
storing a set of measured CRT color data in an SA matrix representing a
relationship between the RGB signal values driving the CRT and a set of
CIE tri-stimulus color values;
setting the RGB signals to a maximum value to display a standardized white
on the CRT;
storing a measured peak intensity vector P relating the RGB maximum
signal values to a set of corresponding CRT intensity values of the
standardized white displayed on the CRT;
subjecting the CRT display to a period of use likely to cause a measurable
change in the standardized white displayed in response to the maximum RGB
signal values;
following the period of use, stepping the RGB signal values through a group
of values substantially spanning from a zero RGB signal value to the
maximum RGB signal values;
storing in the lookup table a sensed CRT intensity value corresponding to
each stepped RGB signal value;
finding in the lookup table a set of maximum CRT intensity values
associated with the maximum RGB signal values;
comparing the maximum CRT intensity values with the peak intensity vector
P , and generating a correction factor intensity vector Pc ;
multiplying the SA matrix by the correction factor intensity vector Pc
to generate a corrected matrix A relating CIE tri-stimulus XYZ color
values to associated CRT intensity values;
receiving a CIE tri-stimulus XYZ color value signal;
dividing the XYZ color value signal by A to generate a corresponding CRT
intensity value;
finding in the lookup table the RGB signal value corresponding to the CRT
intensity value; and
driving the CRT with the corresponding RGB signal value to display on the
CRT the perceptually accurate color in response to the received CIE
tri-stimulus XYZ color value signal.
10. The method of claim 9 further comprising the steps of:
deleting from the lookup table each RGB signal value for which an
associated sensed CRT intensity value is substantially zero; and
inserting in the lookup table CRT intensity values corresponding to any RGB
signal values not deleted or in the group of stepped RGB signal values.
11. The method of claim 10 in which the inserted CRT intensity values are
determined by using a polynomial-based curve fitting calculation.
12. Apparatus for periodically calibrating a cathode-ray tube display (CRT)
to establish a stored relationship between a set of RGB signal values and
associated CRT intensity values such that perceptually accurate colors are
displayed on the CRT in response to color input signal values, comprising:
a CRT driver causing the CRT to display light intensity levels in response
to RGB signal values, a processor generating an initial RGB signal value
for displaying a predetermined color on the CRT, and a memory storing a
lookup table and a measured phosphor intensity vector associated with the
initial RGB signal value;
the processor generating a group of RGB signal values such that a
corresponding group of CRT intensity levels is displayed on the CRT; and
a sensor generating a phosphor intensity vector for each light intensity
level sensed on the CRT, the processor storing in the lookup table the
phosphor intensity vector associated with each of the group of RGB signal
values, finding in the lookup table a current phosphor intensity vector
associated with the initial RGB signal value, comparing the current CRT
intensity vector with the initial CRT intensity vector to generate a
correction factor vector, and altering the color input signal values with
the correction factor to generate corrected RGB signal values that are
sent to the CRT driver for displaying perceptually accurate colors on the
CRT in response to the color input signal values.
13. The apparatus of claim 12 further comprising: a light shield to prevent
ambient light from affecting CRT light intensity level values sensed by
the sensor.
14. The apparatus of claim 12 in which the memory includes an SA matrix
that represents an initially measured relationship between the RGB signal
values driving the CRT and a set of device independent color values.
15. The apparatus of claim 14 further comprising:
a color corrector in which the processor multiplies the SA matrix by the
correction factor to generate a corrected A matrix that relates device
independent color values to associated CRT intensity values, divides a
particular device independent color value by A to generate a particular
CRT intensity value, locates in the lookup table a particular RGB signal
value corresponding to the particular CRT intensity value, and sends to
the CRT driver the particular RGB signal value, thereby displaying on the
CRT a colorimetrically accurate color in response to the particular device
independent color value.
16. The method of claim 15 in which the device independent color value is a
CIE tri-stimulus XYZ value.
17. Apparatus for periodically calibrating a cathode-ray tube display (CRT)
to establish a stored relationship between a set of RGB signal values and
associated CRT intensity values such that colorimetrically accurate colors
are displayed on the CRT in response to CIE tri-stimulus XYZ color input
signal values, comprising:
a CRT driver causing the CRT to display light intensity levels in response
to the RGB signal values, a processor generating a maximum RGB signal
value for displaying an initial white color on the CRT, and a memory
storing a lookup table, a CRT phosphor intensity vector associated with
the initial white color, and an SA matrix representing an initially
measured relationship between the RGB signal values driving the CRT and a
set of CIE tri-stimulus XYZ color values;
the processor generating a group of RGB signal values such that a
corresponding group of CRT intensity levels is displayed on the CRT; and
a sensor generating a phosphor intensity vector in response to each light
intensity level sensed on the CRT, and the processor storing in the lookup
table a phosphor intensity vector associated with each value in the group
of RGB signal values and inserting computed phosphor intensity vectors
associated with any RGB signal values not in the group of RGB signal
values, finding in the lookup table a current phosphor intensity vector
associated with the maximum RGB signal value, comparing the current CRT
intensity vector with the intensity vector associated with the initial
white color and generating a correction factor vector, multiplying the SA
matrix by the correction factor to generate a corrected A matrix
relating the CIE tri-stimulus XYZ color values to associated CRT intensity
values, dividing a particular device independent color value by A to
generate a particular CRT intensity value, locating in the lookup table a
particular RGB signal value corresponding to the particular CRT intensity
value, and sending to the CRT driver the particular RGB signal value,
thereby displaying on the CRT a colorimetrically accurate color associated
with the particular CIE tri-stimulus XYZ color value. |
|
 |
|
 |
Claims |
|
|
|
Description |
|
|
TECHNICAL FIELD
This invention pertains to a system for accurately controlling video
monitor color characteristics.
BACKGROUND INFORMATION
Video monitors comprising cathode-ray tubes (CRT's) are widely used display
devices. Color CRT displays are especially useful means for conveying
graphic information. Generally, a color CRT display includes three
electron guns, each gun controllable to excite an associated red, green or
blue phosphor set carried on the CRT screen. In some applications, such as
graphic arts, advertising, textile design, etc., color CRT displays are
relied upon to display colors having specific colorimetric parameters, so
that the identical color can be reproduced with other media (paints, inks,
dyes, etc.) using those same colorimetric parameters. Such parameters may
be the well-known XYZ tristimulus values as defined by the International
Commission on Illumination or CIE.
Disclosed in a copending patent application of Murch, et al., entitled,
"Display-Based Color System", is a useful system for organizing colors
into a perceptually uniform color space having coordinates that are
related to the CIE tristimulus values XYZ. Also disclosed is a method for
transforming a point in that color space (as defined by the color space
coordinates) into suitable notation for displaying that point (i.e.,
color) on a CRT display. To this end, the point is transformed into a
corresponding rgb intensity vector. An rgb intensity vector is a 3-element
vector denoting the relative intensity contribution of each CRT phosphor
set required for displaying a selected color. To display a color
represented by an rgb intensity vector, it is necessary to convert the rgb
intensity vector into associated "DAC values." DAC values are scaled
numerical values (usually ranging from 0-255) corresponding to the
electron gun control levels required to drive the associated phosphor set
at various luminous intensities.
In order to utilize rgb intensity vectors, or for otherwise producing
colorimetrically accurate displays, the CRT display must be calibrated.
That is, the DAC-value/phosphor intensity relationship must be precisely
determined.
SUMMARY OF THE INVENTION
This invention is directed in part to a system for calibrating the
intensity response of CRT monitor phosphor sets to each of a plurality of
electron gun control levels. Specifically provided is a processor that
generates a sequence of discrete DAC signals. Each DAC signal identifies
an electron gun and a DAC value for driving the electron gun at a selected
control level. Also provided is a monitor driver that is connected to the
processor and to the monitor and is controllable, in response to a DAC
signal, for driving the electron gun identified by the DAC signal at the
control level that is identified by the DAC signal. The driven electron
gun excites an associated phosphor set. A sensor is provided for detecting
the luminous intensity level of the phosphor set excited by the electron
gun, and for converting the detected intensity into a representative
signal. The processor receives the representative signal produced by the
sensor and stores, as intensity data, each DAC value along with the
corresponding phosphor set intensity level.
As another aspect of this invention, the processor processes the intensity
data to generate a functional approximation thereof for constructing
look-up tables defining intensity levels for each of the plurality of DAC
values.
As another aspect of this invention, the system is operable for utilizing
the detected luminous intensity levels for generating matrices suitable
for reversibly transforming rgb intensity vectors into corresponding XYZ
tristimulus values.
As another aspect of this invention, the video monitor control system is
operable for accurately reproducing a color sample on the video monitor.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a diagram of the system of the present invention.
FIG. 2 is a flow chart depicting the sequence of events for calibrating a
video monitor in accordance with the invention.
FIG. 3 is a flow chart depicting the sequence of events for generating
matrices suitable for reversibly transforming rgb intensity vectors into
corresponding XYZ tristimulus values.
FIG. 4 is a diagram of an alternative embodiment of the system useful for
detecting colorimetric data from a color sample and for accurately
duplicating the color on a video monitor.
FIG. 5 is a flow chart depicting the sequence of events for detecting
colorimetric data from a color sample and for duplicating the color on a
video monitor.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
Referring to FIG. 1, a preferred embodiment of the present invention is
employed for calibrating a video monitor 20. The monitor 20 to be
calibrated may be any color CRT device. As is known, the monitor includes
three electron guns, each gun corresponding to a red, green or blue
phosphor set carried on the CRT screen 21.
Each electron gun delivers a beam of electrons to the screen to excite the
associated phosphor set. The phosphor set glows at a luminous intensity
level that corresponds to the beam current or control level of the gun.
The beam is raster-scanned over the screen in response to conventional
beam deflection mechanisms. The monitor 20 is driven by a display
generator system or monitor driver 22. The monitor and monitor driver can
be any suitable color graphics system such as a Tektronix 4125 Color
Graphics Workstation. Further, although the discussion hereafter
identifies a separate main computer 24, it is contemplated that all of the
described functions of the computer may be carried out by the processor
sets incorporated into the Color Graphics Workstation. As a result, such a
workstation will have self-calibration capability. It is also pointed out
that the video monitor color control system of the present invention is
useful for calibrating any personal computer monitor driven with a color
graphics display generator.
The monitor driver 22 receives control signals from the computer 24 via a
conductor 26. The control signals, hereafter referred to as "DAC signals",
are coded digital signals identifying a control level for driving a
particular electron gun. The monitor driver decodes the signals and drives
the gun accordingly.
The main computer 24, such as a Compaq personal computer manufactured by
Compaq Computer Company, includes a processor that is programmed to
generate a sequence of DAC signals, and to supply those signals to the
monitor driver 22. Specifically, as mentioned above, each DAC signal sent
to the monitor identifies an electron gun to be addressed by the monitor
driver, and indicates the DAC value corresponding to a particular gun
control level. Typically, DAC values range in integer increments from 0 to
255.
For each DAC signal, the monitor driver 22 directs the monitor 20 to
produce a full screen display of the color that results when the signaled
electron gun is driven at the signaled DAC value. As each distinct DAC
signal is sent to the monitor, the computer 24 signals a sensor 28 to
supply the computer with a number of luminous intensity level readings
detected from the CRT screen. In this regard, the sensor 28 is any
suitable photometer or radiometer responsive to the monitor screen
luminosity to produce a representative analog electrical signal.
Preferably, the monitor screen 21 is shielded 29 to eliminate ambient
light effects on the sensor.
A sensor controller 30 is interconnected between the sensor 28 and the
computer 24 to facilitate communication therebetween. The sensor
controller includes an analog-to-digital (A/D) converter for reducing the
sensor readings to representative digital signals. Further, the sensor
controller includes a microprocessor such as an Intel 8751, which is
programmed to supply the digital intensity signals to the main computer
via a conductor 32 when signaled by the main computer. Several intensity
level readings are delivered to the computer for each DAC signal. The
computer averages and stores the readings.
After DAC-value/intensity data has been obtained for all three guns, the
data is processed, as described more fully below, to construct a
DAC-value/intensity look-up table for each gun. The look-up tables are
tabulations of every single DAC value along with an associated phosphor
set luminous intensity level. The look-up tables are thereafter available
to convert rgb intensity vectors into the appropriate DAC values, or to
convert the DAC values of a color into its rgb intensity vector.
Turning to FIG. 2, depicted there is a flow chart corresponding to the
computer program used for operating the system of FIG. 1. Specifically the
computer 24 receives as input the DAC value step size, the upper and lower
DAC value limits, and the number of guns. The DAC value step size
indicates the increment between each DAC value sent to the monitor driver
via each DAC signal. Rapid calibration is achieved if the DAC value step
size is 5; however, any step size may be selected.
Lower limit and upper limit DAC value constants are also established.
Specifically, the upper limit DAC value constant is established at 255.
The lower limit constant is established at a value nearest to 0, which
will yield an integer multiple of the DAC value step size between the
upper and lower limits. For instance, for a DAC value step size of 25, the
lower limit is established at 5. Establishing the lower limit as just
described ensures that the intensity level for the highest DAC value, 255,
will always be detected.
The number of guns input to the computer indicates whether a monochrome (1
gun) or color (3 gun) monitor is being calibrated. This discussion assumes
that a color monitor is being calibrated, although monochrome monitors may
be likewise calibrated.
After establishing communications files between the computer 24 and the
monitor driver 22, and between the computer and the sensor controller 30,
the sequence of DAC signals is sent by the computer 24 to the monitor
driver 22. In the preferred embodiment, each DAC signal is transmitted in
a coded form comprising ASCII characters that identify the DAC signal as a
sequence of three elements respectively corresponding to the DAC values
for the electron guns associated with the red, green and blue phosphors.
Two of the three elements are established at DAC values equal to 0, and
the remaining element is the DAC value to be applied to the associated
electron gun.
After each DAC-signal is transmitted to the monitor a "sample" signal is
transmitted to the sensor controller communication file. The sensor
controller reads the "sample" signal and loads the communication file with
a digital intensity level reading that was obtained, via the A/D converter
in the sensor controller 30, from the continuously applied sensor output.
The content of the communication file is read by the computer, which then
signals for another reading. This sampling process is repeated for the
same DAC signal until several (preferably 10) intensity level readings are
obtained and averaged.
The averaged intensity readings are stored as part of an array of
DAC-value/intensity levels. The computer next generates and sends to the
monitor driver another DAC signal corresponding to the next DAC value step
for the same gun signaled in the prior DAC signal. It is pointed out that
the sequence of DAC signals may be sent in ascending or descending order
of DAC values between the upper and lower DAC value limits. It has been
found, however, that phosphor persistence may influence intensity readings
if the monitor is driven with descending DAC value steps. Accordingly, it
is preferred that the DAC signal sequence be sent to the monitor in order
of ascending DAC values.
After intensity levels for all DAC values of all guns have been collected
(the DAC-value/intensity data hereafter referred to as the "intensity
data"), the low end of the intensity data is analyzed to delete that
portion of the data wherein the intensity levels are indistinguishable
from noise. That is, during manufacture and set up of each monitor, the
CRT electron guns are over-biased to ensure that a zero drive voltage
applied to the guns yields no phosphor glow. Thus, it is desirable to
delete the low DAC values having no effect of the phosphor intensity
because of the over-biasing (i.e., the DAC values "producing" only noise).
To this end, the intensity levels of the two lowest DAC values are
compared. If the difference in the levels does not exceed a threshold
value t, the lower level is discarded and the remaining level is then
compared with the next higher level. This comparison process continues for
increasing DAC values until the difference between compared intensity
levels exceeds the threshold t, at which point the intensity data will be
free of all meaningless data.
The threshold value, t, is established as the detectable difference in the
power or intensity function of the CRT corresponding to the change in DAC
values (i.e., gun control levels) between each DAC signal.
The intensity data, comprising as it does a set of ordered pairs, is
amenable to curve fitting techniques for providing a simple analytical
expression of the DAC-value/intensity relationship. Preferably, the data
is processed by the technique known as the least-squares fit method,
wherein the coefficients of the polynominal function
y=a.sub.0 +a.sub.1 x+a.sub.2 x.sup.2 +. . . +a.sub.m x.sup.m(1)
are obtained by solving the system of m+1 equations:
##EQU1##
It has been found that acceptable curve fitting results are obtained when
the degree, m, of the polynomial is between 5 and 7.
Using the calculated coefficients of the polynominal function of equation
(1), a look-up table for each electron gun is next generated and stored by
the computer. Specifically, the look-up table is generated by solving the
polynominal function for each single DAC value between upper limit (255)
and the lower limit remaining after the noise data deletions are made.
Usually, the remaining lower limit DAC value will be in the range of
25-30. It is important to note that the look-up table comprises intensity
levels for every single DAC value, and not just for the DAC value steps
for which the intensity data was originally collected.
It is contemplated that the look-up tables could be constructed without
processing the intensity data as just described. For instance, useful
look-up tables may be constructed by acquiring intensity level readings
for every DAC-value (i.e., DAC value step size=1). However, greater
precision is found with the data processing technique just described.
As another aspect of this invention, the system is readily adaptable to
generate a matrix useful for precisely converting rgb intensity vectors
into corresponding CIE tristimulus values XYZ. That matrix is hereafter
referred to as the RGB-to-XYZ matrix and denoted [A]. Similarly, the
system also generates a matrix known as the XYZ-to-RGB matrix for
converting the XYZ tristimulus values of a color into a corresponding rgb
intensity vector. The XYZ-to-RGB matrix is denoted [A.sup.-1 ].
With reference to FIGS. 1 and 3, to construct the RGB-to-XYZ matrix, the
computer is supplied with two sets of data, one data set is referred to as
the specified RGB-to-XYZ matrix and denoted [SA]. This data is generated
through spectroradiographic analysis of the monitor at the time of
manufacture. The specified RGB-to-XYZ matrix is a 3-by-3 matrix, the
elements of which represent the relative intensity contributions of the
red, green and blue phosphors (columns) to the XYZ tristimulus values
(rows) determined when the phosphors are excited to full intensity. A
useful explanation of the CIE system and the related tristimulus values
XYZ is provided in "Principles of Color Technology", 2nd ed. 1981, by
Billmeyer and Saltzman.
The second data set supplied to the computer is a three element vector
denoted [P], the elements of which represent the peak luminous intensity
level of the red, green and blue phosphors as detected at the time the
specified RGB-to-XYZ matrix data was compiled.
As an important aspect of this invention, the system processes the input
data in conjunction with intensity level readings to generate a corrected
RGB-to-XYZ matrix, which reflects the true phosphor luminance
characteristics of the monitor, as opposed to those characteristics
specified by the manufacturer. In this regard, it can be appreciated that
heavy use, vibrations, etc., can vary the luminous characteristics of a
particular monitor relative to the specified characteristics. Accordingly,
employing the system of the present invention to account for these
variations yields a "corrected" RGB-to-XYZ matrix. This corrected matrix,
denoted [A], is important for generating colors on the monitor that have
the colorimetric parameters desired.
Corrected RGB-to-XYZ matrix construction commences with the computer 24
generating a series of three DAC signals. The first DAC signal instructs
the monitor driver to drive the "red" electron gun to produce a peak
intensity red phosphor glow. That is, the red electron gun is driven at
the maximum DAC value, 255. After the first DAC signal is transmitted to
the monitor driver, the intensity level reading is sampled from the
monitor screen and stored in the computer as described earlier.
The second and third DAC signals similarly signal peak intensity blue
phosphor glow and peak green phosphor glow, respectively. As above, the
peak intensity level readings for these phosphors are sampled and stored.
It can be appreciated that the peak intensity levels may also be obtained
by reference to the look-up tables described earlier.
The sampled peak intensity levels for each phosphor is compared with (i.e.,
divided into) its counterpart of the specified vector [P]. This comparison
yields a correction factor vector [P.sub.c ], the three elements of which
represent, in decimal form, the variation of each actual phosphor peak
intensity level from that specified. Accordingly, when the specified
RGB-to-XYZ matrix is multiplied by the correction factor vector [P.sub.c
], the corrected RGB-to-XYZ matrix is formed.
As noted earlier, the corrected RGB-to-XYZ matrix is useful for
transforming an rgb intensity vector into its corresponding XYZ
tristimulus values.
The corrected RGB-to-XYZ matrix is next inverted to yield the 3-by-3
XYZ-to-RGB matrix, [A.sup.-1 ]. As noted earlier, the XYZ-to-RGB matrix is
useful for transforming the tristimulus values of a color into its
corresponding rgb intensity vector for the CRT.
As another aspect of this invention, the video monitor color control system
is readily adaptable for accurately reproducing a color sample on a video
monitor. With reference to FIGS. 4 and 5, and as described in more detail
below, the sensor 28 and a color filter wheel 40 are employed to detect
the luminous characteristics of a uniform color sample 42. These
characteristics are transformed, using an XYZ-to-RGB matrix, into an rgb
intensity vector for the video monitor. The rgb intensity vector is
converted via the above-described look-up tables into the corresponding
DAC values and the sampled color is displayed on the monitor screen 21.
Turning to the particulars of the color duplication process just outlined,
the color wheel 40 is used to provide the sensor 28 with the luminance
levels of the red, green and blue light that is reflected from the color
sample 42. The light passing through the red, green and blue filters will
hereafter be referred to as the red, green and blue primaries. The color
wheel is a circular arrangement of three color filters and one clear
space. The color wheel is selectively rotatable by known means. It is
contemplated that other mechanisms for interchanging color filters may be
substituted for the color wheel.
The color wheel 40 is analyzed with conventional spectrophotometric means
to establish the colorimetric properties of each filter. In this regard,
the red, green and blue filters are each illuminated by a standard CIE
illuminant and the primary light transmitted therethrough is analyzed to
determine the relative contribution of each primary to the CIE XYZ
tristimulus values of the illuminant. This data is assembled into a color
filters matrix denoted [F]. The color filters matrix is a 3-by-3 matrix
with elements representing the relative intensity contributions of the
red, green and blue primaries (columns) to the XYZ values (rows) of the
illuminant.
Preferably, the color filters matrix is normalized so that the Y
tristimulus value of the illuminant will equal 1. As noted in the
referenced patent application by Murch, et al., this normalization is
useful for correlating the Y tristimulus value with a color parameter of
the CIE system known as the metric lightness function. The metric
lightness function is a useful means of describing the lightness of a
color.
The color filters matrix data is next processed to account for differences
between the illuminant used to establish the matrix and the "reference
white" light, which is the light used for illuminating the color sample.
To th | | |
