JPH025683A - Method of improving printed appearance of computer-formed color image - Google Patents

Abstract

PURPOSE: To obtain a printer data necessary for printing out a subjectively allowable color reproduced image by serially writing/reading out R, G and B items and C, M and Y items or M, Y and K items to/from a printer pixel value retrieving table. CONSTITUTION: An indexed L* retrieving table is constituted of R, G and B monitor pixel values and an optional R, G or B value supplied from a frame buffer 17 is converted into a corresponding L*R, L*G or L*B value based on the table 18. The table 18 applies an L* value suitable for directly converting monitor L*R, L*G and L*B values into corresponding L*C, L*M and L*Y values necessary for reproducing achromatic tones by a printing process 12. Namely the tone reproducing characteristic of the process 12 is indendent of any color correction applied by a color correction stage 15.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention This invention relates to the printing of computer-generated color images, and more particularly to methods for improving the print appearance of such images.

BACKGROUND OF THE INVENTION Modern personal computer systems are often equipped with color monitors and color printers to allow users to view and print color images. However, users of these systems are often disappointed by the poor quality of color reproduction exhibited by the actual printed image, especially when careful consideration is given to the coloration of the displayed image. Although it is recognized that data reproduced for a monitor should be converted into a form suitable for a printer, most users of personal computer systems are not skilled at color reproduction, nor are they capable of advanced color measurement. Because the equipment is not readily available, sufficient premises are not in place to produce a satisfactory transformation. In fact, many of these users accept printed color images of inferior quality to perfect graphic art in exchange for a simple process that allows them to print color images perceptually similar to the displayed images.

Other techniques include colorimetric matching and scaling of displayed and printed images, with the theory that colorimetric matching leads to acceptable reproduction of computer-generated color images.
has been proposed to convert monitor data to printer data based on For example, W, F, Schriba (S
Chrieber), “Color Prepress System Using Appearance Variables”, Muan, of Imaging
Technolog, Vol. 17. Nn 4+
August 1986, pp. 200-210, and M. C. Stone et al. [Explanation of color reproduction methods used in this issue of Color Research and Applications, Col.
p 1ication,,V.

1, LL 1986 Special Issue, Background Information on Colorimetric Matching Techniques, pages 583-588. Also, Gordon et al., ``J., Proceedings or the
Technical Association of G
rahic Arts, 1987,
An interesting and more relevant alternative method is proposed, which involves scale adjustments such as - across the monitor and printer in the ``ull'' color space.

However, such scale adjustment makes the brightness, hue, and saturation of the reconstructed image dependent on each other.

SUMMARY OF THE INVENTION It is therefore possible to enable a user of a computer system, particularly a personal computer system, to obtain the printer data necessary to print a subjectively acceptable color reproduction of a computer-generated image. , there is a need for an easier and more flexible conversion of monitor data to printer data.

According to the present invention, (R, G, B) monitor pixel values supplied for displaying a color image on a color monitor are used to reproduce such an image on a color printer. Gray balanced for (C,M,Y) or (C,M,Y,K)
An equivalent grayscale conversion method is provided for converting to printer pixel values.

To implement this conversion method, the monitor/printer combination is linearized in the standard L*% a*, d color space for neutral tones, from monitor black to monitor white for monitors and from monitor white for printers. Equivalent neutral for individual terms of monitor and printer pixel values for all pixel values ranging from printer black to printer white, 'S (ENL Yu'5)
seek.

Furthermore, based on a scaled monotone ma-noping function that converts monitor blanks into printer whites and converts monitor whites into printer whites, the monitor's gray range is converted to the printer's gray L* range. be done. Monitor pixel values are converted to printer pixel values, for example, ENL indexed by the R, G and B terms of all possible monitor pixel values, appropriately scaled and transformed, and (2
) C, M and Y terms of every possible printer pixel value indexed by these ENL, values or C1
For a search table containing terms in M, Y, each R,
It can be converted by serially reading and reading the terms into G, B terms and C, M, Y or C, M, Y. RlG of the monitor pixel value and ENL of the B term (for example, select the ENL value to be mapped to the term).

A constant monotonic mapping function can be used to directly map C, M, Y or C, M, Y, terms for corresponding printer pixel values (using a criterion). Alternatively, the way each monitor pixel value is mapped to each printer pixel value can be modified using an optional gray-balanced color correction matrix without affecting the color tone. , it is possible to adjust the saturation and hue of the print reproduction. The color tone of the print reproduction is determined by the monitor and the printer's L.
The scaling of the displayed image is based on a monotonic mapping function used to convert to the range, and can therefore be adjusted by adjusting that function.

Further features and advantages of the invention will become apparent from the following detailed description with reference to the accompanying drawings.

EXAMPLES Although the present invention will now be described in detail with reference to specific examples, it should be understood that this is not intended to limit the invention to those examples.

On the contrary, variations of the invention limited by the scope of the claims,
All alternatives and equivalents are intended to be encompassed by this invention.

First, to aid in understanding the following discussion, we will briefly define some of the terms used herein and clarify the measurement methods recommended for use in calibrating the conversion method of the present invention.

■Tomonobu■Degai As is well known, color can be described by three variables.

The red, green, and blue light contained in the light emitted by a CRT monitor or emitted by a light source and reflected from an object determines the characteristics of the light that reaches the eye. Three variables are usually sufficient to define one color stimulus, which is called the r-stimulus. International Commission on Illumination (CIE)
A set of variables (L, am, b) of the perceptual color space is established. This is because these three variables correlate better with the human visual system's response to stimuli than other standard specifications. G., Wyszecki et al., Color Ratio Associations and Methods, Constant Data and Formulas, 2nd Edition, John Wiley and
Sone, 2x-York, 1982, 162-16
See page 9.

The response evoked by a stimulus can also be described by three other variables, such as hue, lightness, and saturation. The magnitude of brightness depends on the relative amount of light incident from the object or image. White objects have high brightness, while black objects have no brightness at all.

L is closely correlated with brightness and is usually assigned a vertical axis in three-dimensional displays. Bright colors such as orange have high saturation, while gray has zero saturation. Hue is associated with color names such as purple, yellow, and green.

For purposes of the following discussion, a pixel is defined as the smallest area of a monitor or print that can be independently colored. The conversion from monitor pixel values to printer pixel values is
Determines the amount of ink required to reproduce each red, green, and blue (R, G, B) triplet displayed on the monitor, and this amount is specific to a particular monitor/printer pair. . Unfortunately, there is no simple relationship between the pixel values that drive the red, green, and blue phosphors of a CGT monitor and the pixel values required to produce an acceptable print. In other words, the inks corresponding to the three additive primary colors of CRT, red, green, and blue, are the three subtractive primary colors - cyan (absorber of red), magenta (absorber of green), and yellow (absorber of blue). be.

As will become clear below, the values used here for optional color and saturation correction are expressed as densities. DRID each monitor's red, green and blue density. , and Dll, and each density of cyan, magenta, and yellow 6, Dll
, I, and Dv. The concentration is log. Defined as 1/r. where r represents the reflectance, in other words the ratio of the amount of light reflected from an area to the amount of light reflected from a white surface, an absolute diffuse reflector.

Each color component of a pixel is described by three or more variables called pixel components. Notation [C1M, Y, K)
indicates the value of the printer pixel.

(C, M, Y) is an abbreviation for (C, MSY, O), and C, M
, Y, and K represent the amount of cyan, magenta, yellow, and black pigment, respectively, that the printing system is required to place at a given pixel location on the page. These pixel values range from O to max. within the range of
It is a value that depends on the conversion. For example, if each color component is defined by an 8-bit byte unit, then these are 0 to 2 bits.
It is an integer value in the range of 55. C=M=Y==0 means that no pigment is applied, while C=M=Y==max means that all pigments are applied to the maximum extent. Between 0 and a maximum (eg 225), the printing system responds monotonically.

As is well known, in photography, equivalent achromatic density (Equiva) is used to define the amount of pigment in an image in terms of visible effect.
lent Neutral Dansity (END
)) concept is used. In three-color photography, the END of one pigment is the density that would result if the amounts of the other two pigments were added in just enough amounts to produce a gray color. This concept is useful for analyzing the brightness reproduction characteristics of other reproduction processes, and is therefore widely adopted for that purpose in graphic arts. However, in the following explanation, the lightness reproduction characteristic will be expressed based on a perceptually linear L scale rather than a logarithmic density scale. Therefore, the concept of equivalent neutral density (ENL) is used instead of equivalent neutral density. The shortened expression used here is, in the case of monitor values,...Lyu. ,L. In other words, Lyu8, Lyu. , L8 are the respective magnitudes of the amount of red, green and blue light emitted from a monitor CRT (in terms of perceived visual effects).
m M , Lm v , and Lmb K are each defined in terms of perceived visual effect and the amount of cyan, magenta, yellow, and black ink to be printed on paper or other recording medium.

The notations (R, G, B) indicate monitor pixel values, and L□, L,. , L□ indicate the amount of red, green, and blue, respectively, that the display system is required to convey to the screen. In other words,
Monitor pixel value (R, G, B) = (0,0,0)
means a very dark color close to black (i.e. “monitor black”), while the value (R,0% Bl = (Il
ax Smax % wax ) is a color very close to white (
In other words, it means "monitor white"). On the contrary, the printer pixel values (C, M
, Y, K) = [0,0, O2O3 means "printer black"], while the value (C, M, Y, K) = (ma
x, naaX % lla such that gray on the monitor is effectively reproduced as gray on the print whenever the monitor pixel values (v, v, v) of are mutually equal.

■、iIN's 9th New Year 8th hand There are many instruments that can be used to measure color stimuli.

Equipment for measuring and adjusting CRT monitors ranges from simple vision devices that adjust the monitor's color balance to sophisticated telescopic spectroradiometers that measure the spectral emissions of the monitor's phosphors at each wavelength. It is. Spectrophotometers, colorimeters and densitometers are used to measure the prints. Although each device has its value, this invention can be practiced with two devices; one to measure monitors and one to measure prints.

It is recommended that both of these devices have a spectral response similar to normal human vision as defined by CIE.

More specifically, monitors are generally measured using a color balance meter. For example, the Minolta TV-Color Analyzer TV-2150 can be used to adjust and maintain the ratio of red, green, and blue radiation in a conventional CRT. This type of device can also determine the luminous intensity at various levels of brightness. If comparable equipment is not available, a well-calibrated and maintained monitor can also be used with good results. However, taking instrument measurements increases the level of confidence and is advantageous when troubleshooting.

The second instrument is a colorimeter, such as the Minolta CR231, which measures the color and brightness of the print to calibrate the printing process. The light source, filter and photodetector combination of such an instrument is designed to respond to CIE X, Y, Z, a portion of which is directly converted to Lm, am, by. Although you can use a colorimeter to measure the density of prints, there are times when a densitometer is more convenient. Of the two types of densitometers (broadband and narrowband), the narrowband is preferred.

m,' Process and its Calibration Referring now to the drawings, and in particular to FIG.
R, G, B) pixel values to enable printing a subjectively acceptable reproduction of the computer-generated color image displayed on monitor 11 by a particular printing process 12 (C, M, Y] or (C , M, Y, K) pixel values. Since the printing process 12 is specific to each printer, the terms printing process and printer are used interchangeably herein. can be replaced with

The monitor's (R, G, B) pixel values are converted to the printer's (C
. This is performed between the monitor 11 and the printer 12 by a print server or the like that includes a contrast adjustment table (also not shown). That is, the conversion is specific to a particular monitor/printer combination.

Achieving and maintaining acceptable print quality in printing process 12 requires routine inspection of the prints it produces. As is well known, the color of a print is affected by the lighting where it is viewed, so the light source and viewing conditions under which test and sample prints are examined should be carefully chosen and kept virtually constant. be. Additionally, a file of test prints printed while determining whether the printer 12 is properly calibrated should be saved for comparison with subsequently printed samples if there are any significant differences between the two sets of prints. The system should be recalibrated if the presence of During system calibration, the color of each ink is measured. L″I of printer 12 for one or more pixel values
It is also effective to appropriately measure the intermediate level of brightness in order to determine whether or not there is a shift toward I-response. If such a shift is detected, the system should also be recalibrated. However, on a routine basis, it is generally sufficient to measure only the maximum density of the ink, such as to determine whether the printing process 12 is producing a constant density from day to day.

Of course, when using a new batch of inks, always check the color by measuring the maximum density of each ink with a colorimeter, and calibrate the system if there is a significant change in the density of each ink being used. It should be possible to fix it.

In practice, the black produced by the maximum concentration of cyan, magenta and yellow pigments alone may not be black enough in some prints and may result in an undesirable hue.

Therefore, the addition of black ink often results in significant improvements in the perceived quality of the printed image by providing a lower hue and higher density of black.

Fortunately, as detailed below, the amount of blanks to be printed is based on the amount of black required.
, L1 values can be obtained directly from the three-color (C, M, Y) display by making it a function of the smallest predetermined set of three of the L1 values.

The conversion process involves three basic steps, the latter two of which are optional; (1) areas that do not contain hue (white, gray, and black) on the monitor image have equivalent significant hue on the print; the desired black-to-black and white-to-white monotonic correlation between the gray values on the monitor 11 and those of the image printed by the printer 12, while ensuring that the gray values are reproduced without causing (2) a black printer stage 14 that complements the three-color image, including a blank image (this recalibrates the gray balance and contrast adjustment stage 13 for four-color printing; and (3) adjusting the hue and saturation of the printed image independently of its tone, without affecting the tone reproduction characteristics established in the gray balance and contrast adjustment step 13, Color correction step 15 so that the area corresponds more closely to the monitor image, such that for monitor pixel values (R,G,B) where R=G=B, Lm@=L*a=L's To determine the correct L and response of the monitor 11, the monitor is first adjusted to maximum contrast consistent with good performance.Each time the monitor 11 is used, it is recommended that both brightness and contrast be adjusted to a predetermined level. Preferably, the monitor 11 should also be turned on and stabilized before using it, i.e. before making any measurements. Equal (R, G, B) values for a given pixel should be It shall produce achromatic white, gray, and black at any brightness.However, it is extremely difficult to visually detect changes in monitor color, so monitor 11 is calibrated to keep the monitor color constant. Typically, the monitor 11 is controlled by a display controller 16 in response to (R, G, B) pixel values received from a frame buffer 17 or the like.

It is often difficult to obtain reliable and repeatable measurements from relatively inexpensive color monitors.

Preferably, all measurements are taken within a short period of time, since even a putatively constant gray color can vary considerably over time. Also, all measurements should be made on the same area of the monitor screen, as some monitors measure up to a 60% difference between the center and edge of the screen. Unfortunately, if white is displayed on the monitor immediately before a constant gray, even a constant gray will produce different L values. Therefore, at least 10 to 20 seconds should be allowed for the monitor 11 to settle after each gray level change. Of course, if the total change in gray level is small, its stabilization time can be reduced.

Monitor 11 and measure with the monitor set to black [Olo, 0
It is appropriate to start after stabilizing the settings. First, after recording the Lyumin (monitor black) value with this setting, the monitor 11 is incremented to Lyu1lIax (monitor white) through about 20 achromatic tone steps, and after the display has stabilized in this question-and-answer step. L
Take measurements. This results in the printing of a curve that describes the monitor color as a function of monitor (R, G, B) pixel values for all neutral tones between monitor blank and monitor white, as shown in Figure 2. There is enough data available to do so. This curve, called the monitor linearization curve, is determined by the combination of specific printer pixel values (C,
M, Y) or (C, M, Y, K]. If the luminous intensity Y of the monitor 11 is measured instead of L, the maximum luminous intensity Yu Calculate the value of L from: L"=116'-Ha-Yu ■ Layer τ "(1) The reason for using the symbol for luminosity instead of the standard CIE symbol Y is simply to avoid comprehensibility. This is to avoid it.

Similarly, given the monitor linearization curve in Figure 2,
An L lookup table 18 (FIG. 1) indexed by R, G, and B monitor pixel values can be constructed, thereby allowing any (7) supplied from the frame buffer 17, as shown at 19, to be constructed. RlG or B values, respectively, as shown in 20, l, LyuG, L11
It can be converted to 11 values.

2. Calibrating the Printer - Deriving an Intermediate Gray Balance Table Step 1 in calibrating the printer 12 is to set its three color zero saturation axes so that they are linear, have values, and are gray. do. For this task, the optional blank printer stage 14 is turned off. A series of test charts with different amounts of cyan, magenta and yellow
within a certain range.

are printed in a sufficient number of combinations to obtain a series of nearly neutral tone batches for each value. One of these charts is shown in FIG. 3, with the labels below each three-color batch representing, from left to right, the C, M, and Y pixel values provided to printer 12. In this chart, the C value for cyan increases from left to right, and the M value for magenta increases from bottom to top. Yellow has the same Y value throughout the chart, but is different from chart to chart. Approximately equal spacing between printer black and printer white.

Enough charts are printed to obtain 8 to 10 approximately neutral tone batches with values. A so-called printer gray balance curve is then obtained by plotting the (C, M, Y) pixel values of the neutral color punch against their color values, as shown in FIG.

The gray color of the colorimeter is equal to the white of the recording medium, ao and b”
is defined as That is, in order to simplify the work of defining the ENL gray equilibrium curve of the printer 11, when measuring a recording medium (for example, paper), a colorimeter is used.
It is convenient to set it to read "=0. This makes it easier to identify grays. On each test chart (Figure 3), within approximately one unit of the defined gray point, If the color steps of the test chart are too coarse to provide such a basis, then the gray (C, M, Y) pixel values are found by interpolation. In practice, Additional color patches close to gray may have to be printed to define some grays with the desired accuracy. An initial printer gray equilibrium lookup table 21 (FIG. 1) for converting values 23 to printer (C, M, Y) pixel values 24 is obtained from these three color printer gray equilibrium curves. However, these lookup table values are corrected when the optional black printer step 4 is included in the conversion process.

Looking at FIG. 4 in more detail, the three gray equilibrium curves for color printers are given.

It will be clear that it defines the relative amounts of cyan, magenta and yellow required for a neutral tone having a value. However, if one of the pigments reaches its maximum before the others, as the yellow pigment does in this particular example, it may not be possible to maintain tonal neutrality in the darkest prints. ENL for the printer, this area along the axis (
The gray shaded area between R1 and R2 in FIG. 4) is known as the area of gray imbalance. Doesn't end completely with R2 2
The extension process of the gray equilibrium curve for two colors can be done arbitrarily,
It's a matter of preference, and here both curves extend by a straight line from the last point R2 that can produce a neutral tone to the minimum (darkest) L that the printer 12 needs to achieve, the required point R1. It is. The paper length value R4 is the maximum value that can be requested.

If the maximum achievable trichromatic tone is dark enough, the next two steps complete the calibration procedure.

Alternatively, these steps may be deferred until after blank printer stage 14 is added.

3. Contrast Adjustment - Derivation of Intermediate Contrast Adjustment Table Before printing any test images, the monitor's linearization function (Figure 2) is combined with the printer's gray balance linearization function (Figure 4). , provide a direct mapping from R, G, and B monitor pixel values to C, M, and Y printer pixel values, respectively. Since it is highly unlikely that the minimum and maximum gray scale values on the monitor 11 will match those values on the printer 12, the monitor's values will vary.

To print a range to a printer and monotonically map it to a range,
Additional transformation and scale adjustment steps are performed. The mapping function must be scaled correctly to map monitor white and monitor black to printer white and printer black, respectively.
The user may adjust the way the function changes to adjust the tone reproduction characteristics of the printing process 12.

Although the following assumes that linear tone reproduction is desirable, the present invention allows the user to select and modify the tone reproduction characteristics of the printing process 12 as necessary to meet any special conditions or personal preferences. It should be emphasized that this provides sufficient flexibility to make it possible.

A correctly scaled linear function for mapping the range of the monitor 11 to the range of the printer 12 begins with the intersection of the monitor white L and the printer white L, as shown in the lower left quadrant of FIG. Find the intersection point of monitor black L and printer black L when their values are projected from the orthogonal L axis whose scale is ideally adjusted, and then connect these two intersection points with a straight line, as shown in the figure. Define a contrast adjustment curve such as Although a linear contrast adjustment curve is not required in practicing the invention, it is at least helpful in determining the intersection points needed to construct the foregoing.

This is because the endpoints of the curve must remain fixed to map monitor white and black to printer white and black, respectively.

Given the desired contrast adjustment curve, L-+t
A lookup table 26 (FIG. 1) for converting the values 20, L-a, L-a, L-a, L-a, L-a, L-a, L-a, 20 into a request as shown at 23 and 29, rescaled to the L-range of the printing process. Configurable. After these rescaling adjustments, as seen in the figure.

Each request is a printer request. , L9. and Lyu,
expressed in the form of a value, as shown at 23 if an optional color correction step 15 is used.

C-, L-yu, and L-yu are different from the values, so the values are as shown in 29, R, L-yu. , and LmB. However, even when using the color correction stage 15, the lookup table 18 is still different from the monitor L11R%L. and Lllllll
The value is the amount of L needed to reproduce neutral tones by the printing process 12. It should be understood that this provides suitable L values for direct conversion to , L and L values, respectively. That is, the tonal reproduction characteristics of the printing process 12 are independent of any color correction provided by the color correction stage 15.

4.1 - To summarize the above, Figure 5 shows that the transformation of the present invention affects the monitor's response (Figure 2), the monitor/printer contrast adjustment curve (lower left quadrant of Figure 5), and the printer's gray response. How to combine the equilibrium curves (Figure 4) to calculate the printer's (C,, M,
Y) indicates whether to obtain a single mapping to pixel values. The result is shown in the upper right quadrant of Figure 5, which is plotted by rotating the remaining three quadrants clockwise starting from the (R, G, B) axes to the (C, M, Y) axes. It was done. That is, for a given (R, G, B
) monitor vixel values for each field (R), green (G) and blue (
The term B) is transferred to the L power response characteristic (lower right quadrant) of the monitor, and its ENL value is first obtained. Next, this L value is determined by using the contrast adjustment curve (lower left quadrant).
Monotonically mapped and scaled to the NL range.

This allows the printer's EN for R, G or B terms to
L, values are complementary to cyan (C) and magenta (M) by using the corresponding gray equilibrium curve (upper right quadrant), respectively.
Or the yellow (Y) term is converted, and the resulting additive three primary color [R, G, B] monitor pixel values are:
Three subtractive primaries (C, M,
Y) converted to printer pixel values. The curve shown in the upper right quadrant of Figure 5 is linked to the curves shown in the remaining three quadrants, so the monitor linearization search table 1 in Figure 1
8. By combining the contrast adjustment search table 26 and the gray balance table 21, the monitor response curve, the contrast eye circumference adjustment curve, and the gray balance curve can be calculated using C, M, and Y indexed by R, G, and B, respectively. It should be clear that it can be quantized as just three lookup tables for .

5. System for Tone Scale Tone Referring again to FIG. 1, the monitor linearization provided by look-up table 18, the contrast adjustment provided by look-up table 26, and the printer gray balance provided by look-up table 21 are In order to verify whether or not the tonal reproduction of 12 is scaled substantially linearly with respect to the tones of the image displayed on the monitor 11, for example in terms of L, some of the A test chart (FIG. 7) containing three achromatic color batches is displayed on the monitor 11, and the printer 12
printed by. These 3 with desired L and spacing
Monitor glue to determine the pixel values that result in a color batch (RlG, B).

It is appropriate to use the response (Figure 2). Obviously, if the batches of prints on the test chart are not substantially gray or substantially spaced apart in L, then the system should be recalibrated.

After the tonal reproduction of the three-color conversion has been adjusted to a satisfactory level, it is recommended that some color images be printed. Although these prints may have better detail than prints obtained directly from monitor data, they may still not result in a subjectively satisfactory reproduction of the displayed image. Most likely the reason for this defect is a lack of color saturation (chroma).

B1 Stage - Optional Black Printer Conversion Set 1, tjU However, significant improvements in the appearance of printed images may be achieved simply by incorporating the optional black printer stage 14. [CM, Y) Convert printer pixel values to (C, M, Y, K) values, where K is C,
A term representing the amount of blank ink to be overprinted on the M and Y color separations), and various algorithms exist to adjust (C, M, Y) to compensate for the additional density provided by the black ink. do. However, when calculating the maximum amount of black to be printed Kmax,
You should pay attention. For example, the transfer characteristics of printing process 11 may limit the total amount of pigment that can be deposited on any given area.

This case applies to inkjet printers.

This is because excessive amounts of liquid ink can result in undesirable over-soaking of the paper.

Similarly, prints produced by some thermal transfer printing processes may suffer from a reduction in density as a result of printing blanks over them at some gray levels.

J, A, C, Yule, h 5 ”E
(DM John Wiley and 5ons
, New York, 1967, 2-87, proposes an algorithm that achieves the dual goals of rendering the "areas of gray imbalance" gray and at the same time increasing the maximum density of the printing process 12. FIG. 7 shows the Neel function for the blank printer conversion process 14 (FIG. 1), which describes the effect of the print blank on the perceived brightness of the print. That is, line segment 21 in FIG.
It shows how the L requirement is met by a representative printer 12 when using only cyan, magenta and yellow pigments. In such three-color printing, all print requests below 84 points in which the printing process 12 reaches Cmax, Mmax or Ymax result in the same output response P2.
occurs. However, the desired response is a matching function as shown by line 22. In other words, the curve 23 represents the desired matching function 2
2, L*cz must be printed in addition to the L*H and L*v values and represents the amount of □.

In fact, the black printer curve 23 is extended slightly beyond the gray imbalance region R+Rz to include the overlap zone Rz Rz.

This overlap is achieved by reducing the slope of the black printer curve 23 when black is first introduced into the printing process 12, thereby reducing the gray imbalance region R of the printing process 12.
+ Ensures that sufficient black is available to compensate for Rz, while at the same time making the point at which black is introduced for the first time in the process inconspicuous.

As will be appreciated, incorporating the black printer stage 14 requires the above-described conversion from ENL monitor data to printer data, since the addition of black extends the contrast range of the printing process 12 and changes its gray balance. need to be recalibrated. In other words, the search values in both the contrast adjustment table 26 and the gray balance table 21 need to be readjusted to account for the effects of using the black printer stage 14.

2. In order to find the value of the maximum amount of black ink that should be printed in the contrast table, the value (Ca + ax % Mmax % Ymax
, K), where O<IKI<maximum available pixel value, a series of batches are overprinted with different amounts of black ink. The L values for these four color batches should decrease monotonically as K increases. If this condition is satisfied for all values, then the Kmax term is set to the maximum available pixel value. On the other hand, if there is a significant deviation from a monotonous response, then the L distortion of all remaining batches with a certain black as the boundary is O<K<Kmax
The batches printed using that black monotonically decreasing with respect to are determined, and this determined pixel value (
Cmax% Mmax% Ymax, Kmax
) Check Kmax from the batch. in any case,
Batch [Cmax, Mmax, Y+++ax s
PO can be directly determined from Kmax). Since this batch represents the darkest gray that can be produced by printing process 12, RO=PO is the darkest gray that can be requested from the printer, hence "printer black." Also print the batch [0,0,0,Kmax) and do that.

is measured to determine PI, which is the L value of the maximum amount of K to be printed. Here, the contrast adjustment table 26
However, it becomes possible to recalculate RO and R4 as upper and lower limits, respectively, of L* contrast of the printing process 12.

3. Blank printer line In order to obtain the printer 12's response to the black print on the table,
A range of batches were printed with black ink only, and these black L.

The value versus black printer pixel value is plotted and the 8th
On a black printer as shown. Define characteristic curves. This curve is then quantized to create a lookup table 30 for converting each black request 32 into a corresponding black pixel value 33, as shown at 31.
(Figure 1) is constructed.

4. The black printer curve 23 of the black printer table shown in FIG. 7 is based on the teachings of Neil,
It was obtained directly by taking data from Neal's Blackprinter curve, mapping it into space, fitting a polynomial to the data, and then scaling the curve defined by this polynomial to fit the unit square.

In short, the procedure first yielded a generic black printer curve, as shown in FIG. 9, that could be scaled to match the requirements of a particular printer. The polynomial function describing the particular global black printer curve shown is y
-4,5x3-1.84x2+2.39x.

The generic black printer curve of FIG. 9 bounded by the unit rectangles [0,0], [1, ■] is then linearly scaled to yield the blank printer curve 23 of FIG. and the upper right corner in Figure 7 (R
O, R1) and (R3, R3), respectively. This curve is then completed by horizontally extending it up to R4 by a straight line at level P3 (the paper level), so that all requests greater than R3 (the level where black is introduced) request printing of a blank color component. We guarantee that you will not.

5. Black printer curve Figure 10 is a chain diagram of the black printer curve 23 in Figure 7, including the characteristics of the printer 12, from response to black printing. It shows how the gray calibration curve for A set of three requested (L”c,
34 (Figure 1)
and then the requested value for that particular pixel at 35, and it is appropriate to set it equal to K. As shown in FIG. 1, the blank printer curve 23 (FIG. 7) is quantized to decompose the curve 23 into a lookup table 36 of LIIK print values 33 driven by the requested LmK values 32. , allowing the requested L value to be converted into an appropriate value 32 for the printing process 12. Next, as shown in the figure, the above printer request value 32 is mapped through the black printer linearization table 30 and the four colors (CS
M, Y, K) Find the value 33 in the black term of the printer pixel values.

6. F Color' - Table Correction At this point, the gray equilibrium curves for each of the four C, M, Y and KM materials can be used to produce gray throughout the newly extended concentration range. . But unfortunately,
The combination of the four gray equilibrium curves is now predictable and
It cannot be expected that a response will be given. In other words, as shown in FIG. 11, the addition of black pigment
The color response is often shifted below its desired linear response 22 (FIG. 7) in the vicinity of the gray imbalance region R1-R2 (the L* response of the printing process is shifted to compensate for the lack of pigment in the gray imbalance region). The problem is that we have to shift to at least that level. For this reason, the C, M, and Y gray balance curves for printing process 12 may have to be adjusted to return the printer's four-color response to its target straight line 22 (FIG. 7).

To determine the changed shape of the C, M and Y gray balance curves, a four color test chart as shown in FIG. 12 is printed based on the already existing C, M and Y gray balance curves. Each batch on this test chart recalculates the C, M and Y gray equilibrium curves substantially as described above;
Preferably, it is approximately uniformly distributed in L-space so as to provide updated values for the gray balance lookup table 21 that need to be taken into account when using the black printer stage 14. FIG. 13 shows another Jones (J) conversion process modified to include a black printer stage 14.
This is a chart.

C9th Floor ■ - Side Quality and Color Returning again to Figure 1, a simple 3x3 gray-balanced color correction matrix 37 can be used in an optional color correction step to adjust the saturation and hue of the printed image if desired. To adjust, the mapping from non-chromatic CR, G, B] monitor pixel values to (C, M, Y) or (C, M, Y, K) pixel values may be changed, as shown at 38. In an exemplary color correction procedure,
The input and output are the density vectors, and as shown in 41, the monitor ENL and values [L Yuiri, Iri, B) after contrast adjustment obtained in 29 are combined with the density values (DR, B) appearing in 42.
Dc, Da), and furthermore, as shown in 43, the printer density value after color correction appearing in 44 [DC, Dl,
Dy] is given to 23. Requested printer EN after color correction
We assume that there is an additional mapping step to reconvert,L,to values (L”,,L”,,L”,).Each from,L,to concentration and from concentration to, Conversion 41
A convenient place to perform 43 is after the contrast adjustment process 27 but before the printer's gray balancing 22.

More specifically, the color correction matrix 37 has the following general form; D c"' a ++ DR+ a +zDc + a
13DID m= a zIDR+ a zzD(,
+ a z3D++D ,=a,,D,+a,,D,+
a.3D.

The coefficients a0, a2□, and aff3 are the main elements of this matrix, and the remaining coefficients are the subelements. matrix 37
is the resulting cyan, magenta and yellow density D
Since each phase of the coefficients C, D, and C always has a unit value (that is, DC=D, =D, =1), the gray is balanced. Of course, if the cyan, magenta and yellow inks are exact complements of red, green and blue, respectively, then the color correction matrix is a unit matrix (i.e. its principal elements a1, a2 and a33 are each equal to l, and the remaining subelements are each equal to 0), and no color correction is needed or provided. Unfortunately, such a complementary relationship
This cannot be obtained using known inks.

A trial and error procedure is suitable for adjusting the color correction matrix 37 to allow the user to adjust the saturation and hue of the print to suit individual subjective tastes. Test prints are first generated on the printer 12 using the identity matrix or a matrix adjusted based on previous experience, and these prints are then compared with the corresponding images displayed on the monitor 11. If desired, the overall saturation of such a print can be increased by equally incrementing all major elements of the color correction matrix by small amounts, while decrementing each of its minor elements by half the same amount. Doing the opposite would reduce the overall saturation of the print.

Saturation or unsaturation of a specific hue is determined by 1 of the correction matrix.
This is achieved by selectively varying the coefficients of one or two rows. The table below summarizes the meaning of the subelement adjustments that provide corrections for the six main hues; Excess Red Excess Yellow Excess Green Excess Blue Red N/A
-a (31) N/A + a (
31) Green N/A -a (32) N/
A + a (32) Blue -a (13)
N/A +a(13) N/A cyan
N/A N/A a(31)
a (21) magenta - a (32)
N/A N/A -a (12
)I xu--a (23) N/A -a
(13) N/A However, the matrix coefficients should again be adjusted to give the best coloring of the average image.

These matrix coefficients may be changed many times before the final set is established. However, any of these or other color correction techniques used are at best a best compromise, as generally brighter and more saturated colors can be obtained on the monitor 11 than by the printing process 12. You will find that there is not.

It will also be appreciated that since it is unlikely that any one set of coefficients in color correction matrix 37 will be optimal for all images, it may be desirable to adjust the coefficients for special or particularly important images. Should. - By way of example only, matrix coefficients that have been found to be subjectively sufficient to reproduce a color image that can be considered to be adequately representative of the "average image" are: 1, 2 - 0. 1 -0.1 0.35 1.45 -0.1 -0.30 -0.35 1.65 The above coefficients are useful for determining optimal matrix coefficients for other printing processes and other images. This will be a good starting point.

When using this optional color correction procedure 15, all possible mappings from monitor (R, G, B) pixel values to densities are precomputed and stored in a lookup table driven by monitor pixel values. be able to. We also perform matrix multiplication between a set of established matrix coefficients and their color-corrected densities (De, D
,, Dy] value or request [L"C%L'"ll, L"
,] Mapping to values is also calculated in advance, and scale-adjusted but uncorrected density values (DR, DG, Dll
) can be stored in a search table indexed by That is, color correction can be performed by several table lookups and addition operations.

Conclusion (Effects of the Invention) From the above explanation, the present invention has three additive primary color terms [RlG, B
) by subtractive three primary color terms (C,
4 colors (C
, M, Y, K) with or without adding a blank printer term in the case of printing, and the saturation and/or
It will be apparent that the present invention provides a relatively simple and flexible method for converting with or without hue correction. The conversion process is based on the ENL, response of a particular monitor/printer pair for a given printing process based on a predetermined scaling function. calibrated to match the color tone of the image displayed on the monitor. However, the conversion process may be used with uncalibrated monitors and/or printers if this predictive ability is not needed, as is the case for some users. A user can also change the L* tone scaling for a particular monitor/printer pair by making a few simple adjustments to the way monitor EN values are mapped to printer ENL values. Since the tonal scaling is independent of the saturation and hue of the reconstructed image, users who choose the proposed optional color correction process can perform color correction without the risk of disturbing the tonal scaling of the conversion process. .

[Brief explanation of the drawing]

FIG. 1 is a flowchart illustrating the monitor data to printer data conversion method of the present invention, including an optional blank printer conversion and an optional color correction method; FIG. Figure 3 shows the black and white representation of a typical 3-color test chart; Figure 4 shows the 3-color ENL' gray equilibrium curve for a typical printer; Figure 5 shows the black and white representation of a typical 3-color test chart; Jones of the curve shown
es) Chart chain, including the monitor/printer's inherent contrast adjustment curve; Figure 6 is a three-color gray wedge black and white display of a test batch substantially spaced in Lo; Figure 7 is a three-color gray wedge black and white display; M, Y and 4 colors C, M, Y,
A graphical display showing the correlation between the print request and the print L value; FIG. 8 is a black printer L linearization curve for a typical printer; FIG. 9 is a comprehensive blank printer curve; FIG. In the chain diagram of printer characteristic curves in Figure 8, the black printer curve in Figure 7 is for four colors (C, M,
Y, K) gives the printer's gray balance curve for the black component of the print; Figure 11 shows the effect of a black printer on the correlation function of Figure 7 when only the C, M and Y color components are gray balanced. Figure 12 is a black and white representation of a typical 4-color test chart, such as that used to adjust the 3-color C, M, and Y gray balance curves for 4-color printing; In the Jones chart chain of monitor-response characteristic curves shown in the figure, printer gray balance curves and monitor/printer contrast adjustment curves have been modified for four-color (C, M, Y, K) printing. 11...Color monitor 12...Color printer (printing process), 20...Monitor's characterizing step, 23...Printer's characterizing step, 27...Converting step, 26...Indexed step (contrast adjustment search table). I R2 PRINT ENL''' FIG, 3 O OPEN L' OPEN L''

Claims (1)

[Scope of the Claims] Monitor pixel values expressed in additive three primary color terms for displaying a color image on a color monitor are expressed in subtractive three primary color terms for printing a copy of said image on a color printer. In a method of converting all monitor pixel values into printer pixel values with respect to the additive three primary color terms, by characterizing said monitor with L^* over the entire neutral tone response range extending from monitor black to monitor white. determining the chromatic L^* value; by characterizing the printer with L^* over the entire neutral tone reproduction range ranging from printer black to printer white, the equivalent neutral in terms of subtractive three primary color terms for all printer pixel values; determining the L^* value; converting the characterized response range of the monitor to the characterized response range of the printer based on a monotonic function scaled to map monitor black to printer black and monitor white to printer white; and the equivalent neutral L* values of the additive three primary terms mapped to the equivalent neutral L* values of the subtractive three primary terms;
By indexing the subtractive three-primary terms of the printer pixel values by the additive three-primary terms of the monitor pixel values, the monitor pixel values are tonal scaled by L^* with respect to the responsive image being displayed on the monitor. converting the image reproduction into printer pixel values for printing.