JP2529404B2 - Masking coefficient determination device - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a masking coefficient determination device for determining a masking coefficient for faithful color reproduction in a color printer that prints out a color image.

Conventional technology To perform full-color recording, C (cyan), M
It is necessary to perform gradation recording for each of the three primary color inks (magenta) and Y (yellow). Gradation recording is a density gradation method that can control the density within a single dot, as typified by a sublimation thermal transfer method and a silver salt photographic method, and dither and density methods such as a fusion thermal transfer method and an electrophotographic method. Depending on the pattern, it can be roughly classified into an area gradation method in which gradation is expressed by a combination of dots by utilizing the visual integration effect.

Both methods use the principle of subtractive color mixing using (C, M, Y), which is a complementary color of the additive color mixture of (R, G, B) which is the three primary colors of colored light. With additive color mixing, only the color reproduction range of the three primary colors is a problem, and while each spectral distribution does not affect color reproduction, in subtractive color mixing, the spectral distribution of dyes has a large effect on color reproduction. The actual spectral absorption characteristic of ink is a phenomenon in which the hue of the recorded image changes and the saturation decreases due to the fact that the central wavelength is out of the ideal and the absorption characteristic is broad, so that secondary absorption exists. Occurs.

Conventionally, a technique called masking has been used for these problems mainly in the printing field. The one most often used is the so-called linear masking shown in equation (1). In linear masking, the actual ink density signals (C, M, Y) are the primary color density signals (D _R , D) of the three primary color luminance signals (R, G, B) that are the complementary colors of the three primary color brightness signals (R, G, B) as shown in equation (1). _G , D _B ) matrix operation.

In linear masking, {a _kl } ( _k = 1 in Eq. (1)
˜3, _l = 1 to 3) is called a masking coefficient.

By the way, in linear masking, the additive rule that the sum of the increase in density due to the three dye amounts is equal to the increase in each density component, that is, the additive rule of density in subtractive color mixing (Lambert-Beer rule) is satisfied. That is an implicit assumption. However, in the actual ink, since the additive and additive rules regarding the density are not satisfied, the masking coefficient {a _kl } is often not theoretically obtained. Conventionally, the following two methods are used. It is often used.

The first is determined by repeating the experiment. A color original is color-separated by a scanner to obtain main density signals of the three primary colors. Then, in the printer, a masking calculation is performed on the main density signals of the three primary colors to convert them into density signals of ink, and the density signal is used to control the ink application amount to make a color sample. At this time, an appropriate initial value is set for the masking coefficient in the masking calculation. The obtained color sample and the color original are visually compared or the color signals which are the colorimetric results of both are compared, and the masking coefficient is changed according to the difference between the color sample and the color original. Then, using the changed masking coefficient, similarly create a color sample,
Again compare with the color manuscript. That is, the masking coefficient is determined by repeatedly changing the masking coefficient, creating the color sample, and comparing the color sample with the color original.

The second is a method of executing the least squares method regarding the density. This method will be described with reference to FIG. FIG. 4 is a model of a color reproduction system in which this method is used. X is a known density signal, and a sufficiently large number (n) of X is used to create a color sample by the target printer, and the sample is color separated by a scanner to obtain the three primary color main density signals D. Considering that the transfer function of φ is affected in this process, the color correction system should be provided with this inverse characteristic so that X ′ after passing through the color correction system and the original X should be minimized on average.
^-1 is decided. When the color correction system φ ⁻¹ is realized by the linear masking represented by the equation (1), for example, the known density when a color sample is created by a printer for cyan is
If Cj (j = 1 to n) is set and the result of color separation of the color sample by the scanner is (D _R j, D _G j, D _B j),
A _{11 to} a ₁₃ that minimize the squared error E ² expressed by the equation (2).
Is obtained by the method of least squares.

E ² = Σ {Cj−a ₁₁ D _R j−a ₁₂ D _G j−a ₁₃ D _B j} ² −− (2) For magenta and yellow, the masking coefficient is determined by performing the least squares method in the same manner. This is the method of request (for example, "Image processing for color reproduction", Photographic Industries separate volume "Imaging Part 1").

However, in the first conventional method, the masking coefficient is changed by trial and error, the color sample is created by the recording experiment, and the comparison with the color original is repeated, and the masking coefficient is determined. Requires a lot of time and labor, and it is difficult to judge the optimality for the masking coefficient,
There is a problem that it is very difficult to find a masking coefficient that makes a color original and a color sample of a printer output the same color for a large number of colors.

The second conventional method is a masking coefficient that minimizes the difference between the obtained three primary color main density signals and known densities by color-separating a color sample of a printer output created using known densities with a scanner. Is obtained by a convergence calculation, and it is excellent in that the masking coefficient can be determined by one recording experiment.

However, the obtained masking coefficient will be corrected not only with the spectral absorption characteristics of the ink, but also with the spectral distribution characteristics of the color separation of the scanner, and it will look like a video printer that prints out the color image output to the CRT. However, there is a problem that it cannot be applied to the determination of the masking coefficient of the printer in the system without the scanner.

Further, while the masking calculation should minimize the difference between the color reproduced by the printer and the target color for color reproduction, the masking factor obtained by the second conventional method is
Since the difference between the density signals is minimized and the color difference perceived by humans is not minimized, the correction characteristics by the masking calculation using the obtained masking coefficient are insufficient. Have challenges.

In view of such a point, the present invention can be optimized for a large number of colors without repeating experiments, can be applied to a printer in a system that does not include a scanner, and the determined masking coefficient is perceived by humans. It is an object of the present invention to provide a masking coefficient determination device that can minimize the color difference that occurs.

Means for Solving the Problems In order to solve the above-mentioned problems, the masking coefficient determination device of the present invention has n sets of density signals (Cj, Mj, Yj) (j = 1 to n,
n is a natural number) and a color printer controls the ink application amount based on the density signals (Cj, Mj, Yj) of the color printer to measure n sets of color chips, and color-code color signals Of the density signal (Cj, Mj, Yj) to convert the density signals (Cj, Mj, Yj) into main color density signals of the three primary colors (D _R j, D _G j, D _B j); Calculation of the color difference between the output and the output of the inverse masking calculation means, determination of whether the color difference is the minimum, updating of the inverse masking coefficient according to the determination result, and output of the inverse masking coefficient that minimizes the color difference The control means and the inverse function calculation means for obtaining the inverse function of the inverse masking coefficient output from the control means and calculating the masking coefficient are provided, and the masking coefficient is determined by the convergence calculation.

An action color chart signal generating unit outputs a color chart in which the ink application amount is controlled by a plurality of sets of density signals, and the color measuring unit measures the color chart to obtain the color signal of the color chart. Further, the inverse masking calculation means converts the density signal into three primary color main density signals.

Next, the control means calculates the color difference using the output of the color measurement means and the output of the inverse masking calculation means. When the control means determines that this color difference is not the minimum, the inverse masking coefficient is updated and set in the inverse masking calculation means. The above process is repeated until it is determined that the color difference is the minimum, and the inverse masking coefficient that minimizes the color difference is output.

Further, the inverse function calculating means determines the masking coefficient by obtaining the inverse function of the inverse masking calculation using the inverse masking coefficient that minimizes the color difference between the two color signals.

Example A configuration of an example of the present invention will be described with reference to the drawings. The printer of this embodiment prints out a color image output to a CRT (so-called video printer).
And the linear masking expressed by the equation (1) is used for the masking calculation. Prior to the description of the masking coefficient determination device of the present embodiment, the signal flow of the target video printer will be described. FIG. 2 shows a signal model of the video printer.

The part surrounded by the dotted line in FIG. 2 is the flow of signal processing in the printer, and 21 is the complementary color of the three primary color luminance signals (R, G, B) to the three primary color main density signals (D _R , D _G , D _B ). complementary color conversion means for converting, 22 the output of the complementary color conversion means _{21 (D R, D G,} D B) ink density signals subjected to masking calculated for (C, M, Y) masking calculating means for outputting, 23 Masking calculation means 22
Is a recording control means for controlling the thermal energy applied to the ink in accordance with the output (C, M, Y) of the ink, and 24 is a recording head for applying the thermal energy to the ink to perform gradation color recording. Reference numeral 25 is an output image of the printer obtained by performing gradation color recording by the recording head 24, and the reproduced color signals are (X _I , Y _I , Z _I ). On the other hand, 26 is a CRT, which drives the phosphors of each color with the same signals as the three primary color luminance signals (R, G, B) input to the printer, and outputs color signals (X _O , Y _O , Z _O ). . In the NTSC CRT, the color signal can be obtained by the matrix calculation represented by the equation (3) using the three primary color luminance signals for driving the phosphors.

Complementary color conversion is a three-primary luminance signal (R,
G, B) are the primary density signals (D _R , D _G ,
D _B ), the printer used in this embodiment uses the calculation of equation (4), and the ROM table constitutes the complementary color conversion means 21.

D _R = log (1 / R) D _G = log (1 / G) --- (4) D _B = log (1 / B) The printer used in this embodiment uses masking calculation as
Linear masking expressed by equation (1) is performed,
The masking calculation means 22 outputs the output of the complementary color conversion means 21 and (1)
The masking coefficient {a _kl } ( _k = 1 to 3, _l = 1 shown in the equation
(3) to (3) are performed to obtain the ink density signal. Further, the target value of the color reproduction of the printer used in this embodiment is the output color signals (X _O , Y _O , Z _O ) of the CRT 26, and the masking calculation is the output color signals (X _I , Y _I , Z) of the printer. _I )
The purpose is to make it equal to the output color signals (X _O , Y _O , Z _O ) of the CRT26. Therefore, the masking coefficient {a _kl }
Is the spectral absorption characteristics of the ink actually used to reduce the color difference between the printer output color signals (X _I , Y _I , Z _I ) and the CRT26 output color signals (X _O , Y _O , Z _O ). The center wavelength of is deviated from the ideal, and the absorption characteristic is broad, so that the sub-absorption existing is corrected.

Then, the masking coefficient determination device of this embodiment is deformed a part of the signal model of a video printer of FIG. 2, the concentration signal of the ink (C, M, Y) of three primary colors main density signal from the (D _R, D _G, D
_B ) is an inverse matrix {a ′ _kl } of {a _k1 }, and the matrix operation of equation (5) is used to _{convert the three} primary color main density signals (D _R , D _G , D _B ) to the three primary color luminance signals (R, G, B) is calculated by equation (6), and the printer output color signals (X _I , Y _I , Z _I ) and CRT2
The least squares method concerning the color difference with the output color signals (X _O , Y _O , Z _O ) of 6 is executed to obtain the optimum inverse masking coefficient {a ′ _kl }.

Next, the masking coefficient determination device of this embodiment will be described.

FIG. 1 is a block diagram of the masking coefficient determination device used in this embodiment. In the dotted line in FIG. 1, 8 is a printer used in the present embodiment, and 9 is a color chart created by the printer 8. Reference numeral 1 is a color chart signal generating means for outputting density signals (Cj, Mj, Yj), and 2 is color measurement of a color chart 9 created by the printer 8 to obtain color signals (X _I j, Y _I j, Z _I j). The color measuring means 3 for outputting outputs the color chart signal generating means 1 (Cj, Mj, Yj) and the inverse masking coefficient {a ' _kl } ( _k
= 1 to 3 and _l = 1 to 3) and inverse masking calculation means 4 for outputting the three primary color main density signals (D _R j, D _G j, D _B j) and inverse masking calculation means 3 Output (D _R j, D
Inverse complementary color calculating means for performing inverse complementary color calculation represented by the equation (6) on _G j, D _B j) and outputting three primary color luminance signals (Rj, Gj, Bj),
Reference numeral 5 represents the output (Rj, Gj, Bj) of the complementary color calculation means 4 into a CRT output color signal (X _O j, Y _O j, Z _O j) by the NTSC CRT color conversion formula represented by the formula (3). Color conversion means for conversion, 6 is an inverse masking coefficient {a ' _kl } for the inverse masking calculation means 3.
Initial value setting, output of colorimetric means 2 (X _I j, Y _I j, Z _I j)
Calculation of the color difference between the color conversion means 5 and the output (X _O j, Y _O j, Z _O j) of the color conversion unit 5, determination of whether the color difference is minimum, and updating of the inverse masking coefficient {a ′ _kl } according to the determination result. , Or a control means for outputting an inverse masking coefficient {a ′ _minkl } that minimizes the color difference, and 7 is an inverse function calculating means for outputting a masking coefficient by calculating an inverse matrix of the inverse masking coefficient output from the control means 6. Is.

In this embodiment, the color chart signal generating means 1, the inverse masking calculating means 3, the inverse complementary color calculating means 4, the color converting means 5, the controlling means 6 are provided.
The inverse function calculating means 7 constitutes each means as firmware of one electronic computer. Also,
Σ80 (manufactured by Nippon Denshoku Industries Co., Ltd.) is used as the colorimetric means 2, and the output of the colorimetric means 2 is input to the electronic computer off-line.

The printer 8 is internally provided with the complementary color conversion means 21 shown in FIG.
Masking calculation means 22, recording control means 23, recording head 24
However, when determining the masking coefficient in this embodiment, the masking coefficient of the masking calculating means 23 is set to (3 × 3) so that the contents of the ROM table of the complementary color converting means 21 become equal to the output. Are changed to the elements of the diagonal matrix of. By this change, the printer 8 controls the ink application amount by the output (Cj, Mj, Yj) of the color chart signal generating means 1,
The color chart 9 will be created. In this embodiment, the density from the paper surface density of each color ink of cyan, magenta, and yellow to the maximum density that can be reproduced by the printer is divided into 4 equal parts for cyan and magenta, and 6 equal parts for yellow. The signal was generated by the color chart signal generating means 1.

In this embodiment, color signals (X _I j, Y _I j, Z _I j) and (X _O j, Y _O
As the color difference of j, Z _O j), the distance between two points in the uniform color space converted so that the color difference is perceived evenly is adopted in order to approximate the color difference perceived by humans. That is, the control means 6 controls (X _I j, Y _I j, Z _I j) and (X _O j, Y _O j, Z).
_O j) is converted into coordinates (L ^* , u ^* , v ^* ) in the uniform color space by the expression (7), and the average value of the distances in the n sets of uniform color spaces represented by the expression (8). to update the inverse masking coefficients {a _'kl} to minimize Euv.

^{L * = 116 (Y / Y} n) ^ (1/3) -16 (Y / Y n> 0.008856) 903.29 (Y / Y n) (Y / Y n ≦ 0.008856) u * = 13L * (u'- u _n ′) v ^* = 13L ^* (v'-v _n ′) u '= 4X / (X + 15Y + 3Z) v' = 9Y / (X + 15Y + 3Z)-(7) However, the standard light source used for illumination is the C light source. In the case of 2-degree visual field, Y _n = 100 u _n ′ = 0.2009 v _n ′ = 0.4609 Euv = (1 / n) Σ {(L ^* _O j-L ^* _I j) ² + (u ^* _O j-u
_{^{* I j) 2 + (v}} * O j-v * I j) 2} 1/2 - (8) will be described an operation of masking factor determination device in the present embodiment with reference to FIG. 3. FIG. 3 is a flow chart showing the operation. The operation will be described below step by step.

First, the color chart signal generating means 1 generates a density signal (Cj, Mj, Yj) (step S1), and the printer 8 causes the density signal (Cj, Mj,
Yj) is used to control the amount of ink applied for each color to create 96 color patches (step S2), and the colorimetric means 2 measures the colors of the created color patches to obtain color signals (X _I j , Y _I j, Z _I j) is output (step S3), and the control means 6 calculates the color signal (X
_I j, Y _I j, Z _I j) are coordinates (L ^* _I j, u ^* _I j, v ^* ) in the uniform color space ^.
_I j) (step S4).

On the other hand, the control means 6 sets the initial value {a ' _kl } of the inverse masking coefficient to the inverse masking calculation means 3 (step S5), and the inverse masking calculation means 3 uses the inverse masking coefficient {a' _kl }. The density signal (Cj, Mj, Y
j) is converted into the three primary color main density signals (D _R j, D _G j, D _B j) (step S6), and the inverse complementary color conversion means 4 uses the formula (6) to calculate the three primary color main density signals (D _R j, D _G). j, D _B j) are the three primary color luminance signals (Rj, G
j, Bj) (step S7), and the CRT color conversion means 5 converts the three primary color luminance signals (Rj, Gj, Bj) into CRT output color signals (X _O j, X _O j, by the NTSC color conversion formula (3). Y _O j, Z _O j) (step S8), and the control means 6 uses the equation (7) as in step S4 to obtain the CRT output color signal (X _O j, Y _O j, Z _O j) with uniform color. converting the coordinates of the spatial ^{_{(L * O j, u *}} O j, v * O j) (step S9).

The control means 6 is obtained in steps S4 and S9 (L
^* _I j, u ^* _I j, v ^* _I j) and (L ^* _O j, u ^* _O j, v ^* _O j) are used to calculate the color difference Euv represented by the equation (8) (step S1).
0).

Subsequently, the control means 6 determines whether the Euv obtained in step S10 is the minimum (step S11), and if it is not the minimum, the inverse masking coefficient { _a'kl } is updated and set in the inverse masking calculation means 3. Yes (step S12).

Then, steps S6 to S11 are executed using the newly set inverse masking coefficient. Eu in step S11
The above calculation loop is repeated until it is determined that v is the minimum, and when it is determined that Euv is the minimum, the calculation loop is exited, and the control means 6 sets the inverse masking coefficient {a ′ _minkl } that _{minimizes Euv} . Output.

The inverse function calculating means 7 obtains the masking coefficient {a _kl } from the inverse masking coefficient {a ' _minkl } obtained above (step S13). Since linear masking is used in this embodiment, the masking coefficient is obtained as an inverse matrix of the inverse masking coefficient.

It is well known that the least squares method for the color difference Euv executed by the control means 6 is not linear, but can be solved by a numerical iterative solution method by a non-linear mathematical programming method called an optimization method. In this embodiment, the Fletcher Powell method is used. The optimization method was used.

An example of the masking coefficient obtained in this embodiment is shown in equation (9).

In this embodiment, the printer to which the masking coefficient is applied is a video printer, and the three primary color luminance signals (R, G, B) are used.
The color conversion means for converting the color signals into the color signals (X _O , Y _O , Z _O ) is the NTSC CRT, and the color conversion formula is the formula (3). However, the masking coefficient of the printer in the copying machine is determined. In this case, the color conversion means is a color scanner, and the color conversion formula is such that the color signal (X _O ,
By using the inverse function of the function for converting Y _O , Z _O ) into the three primary color luminance signals (R, G, B), the masking coefficient in the printer of the copying machine can be determined in the same manner as in this embodiment.

Further, in the present embodiment, the CRT output color signals (X _O j, Y _O j, Z
_O j) and the output color signals (X _I j, Y _I j, Z _I j) of the printer were converted into the coordinates of the L ^* u ^* v ^* system uniform color space.
(10) each ^L * ^a * ^{b *} system coordinates of a uniform color space by formula ^{_{(L * O j, u *}} O j, v * O j), (L * I j, u * I j, v * I j)
It is clear that the same effect can be obtained even if the color difference according to the equation (11) is minimized by converting to and.

^{L * = 116 (Y / Y} n) ^ (1/3) -16 (Y / Y n> 0.008856) 903.29 (Y / Y n) (Y / Y n ≦ 0.008856) a * = 500 {(X / X _{n) ^ (1/3) - (} Y / Y n) ^ (1/3)} b * = 200 {(Y / Y n) ^ (1/3) - (Z / Z n) ^ (1 / 3)} −− (10) However, when the standard light source used for illumination is the C light source and the field of view is 2 degrees, X _n = 98.072 Y _n = 100.000 Z _n = 118.225 Eab = (1 / n) Σ {(L _{^{* O j-L * I j}} ) 2 + (a * O j-a
_{^{* I j) 2 + (b}} * O j-b * I j) 2} 1/2 - (11) In addition, tristimulus values the color difference _{(X O j, Y O j} , Z O j) and (X _I j, Y _I
j, Z _I j) or the luminance signal (R _I
j, G _I j, B _I j) is output (R _I j, G _I j, B
_I j) and the output of the inverse complementary color conversion means (Rj, Gj, Bj), and the colorimetric means that outputs the density signal (C _I j, M _I j, Y _I j) is used (C _I j, M _I j, Y _I j) and the output of the inverse masking calculation means (D _R j, D _G j, D _B j) to determine the masking coefficient determined by the printer in the color reproduction. Although the color difference between the color to be reproduced and the color reproduced by the printer does not seem to be a minimum for human eyes, the structure of the device is simplified and the masking coefficient can be determined at high speed.

Further, in the present embodiment, the printer that uses thermal energy has been described, but it is obvious that the difference in the recording principle of the printer is irrelevant to the present invention.

According to the present invention, the masking coefficient used in the masking calculation for determining the density signal of the actual ink from the primary density signals of the three primary colors that are the complementary colors of the luminance signal can be calculated without repeating the color chart creation and color measurement by the printer. Since it can be determined by the convergence calculation, not only can it be determined in a short time, but also the optimum value can be obtained.

In addition, CR is used as the conversion formula from the three primary color luminance signals to the color signals.
By using the color conversion formula of T, it can be applied to the determination of the masking coefficient of a printer in a system without a color scanner such as a video printer.

Moreover, in the present invention, since the error evaluation is performed using the color difference instead of the density signal, the color difference can be the distance between two points in the uniform color space, and the determined masking coefficient is
In the color reproduction by the printer, the color difference between the target color and the reproduced color of the printer is made to be the smallest seen by humans, and extremely faithful color reproduction is possible.

[Brief description of drawings]

FIG. 1 is a block diagram of a masking coefficient determination device according to an embodiment of the present invention, FIG. 2 is a signal model diagram showing a signal flow of a video printer in the same embodiment, and FIG. 3 is a view of an embodiment of the present invention. FIG. 4 is a signal model diagram for explaining a conventional example, and FIG. 4 is a flow chart showing the operation of the masking determination device. 1 ... Color chip signal generation means, 2 ... color measurement means, 3 ... reverse masking calculation means, 4 ... reverse complementary color conversion means, 5 ... CRT
Color conversion means, 6 ... control means, 7 ... inverse function calculation means,
8: color printer, 9: color chart.

Claims (3)

(57) [Claims] 1. A subtractive color primary density signal (D _R , D _G , D _B ) of a subtractive color mixture obtained from a luminance signal (R, G, B) of an additive color mixture
The ink density signals (C, M,
Y) to determine the masking coefficient to be used in the masking calculation, and n sets of density signals (Cj, Mj, Y
j) (j = 1 to n, n is a natural number), and the color printer generates the density signals (Cj, Mj, Yj).
Controls ink coating amount colorimetry n sets of color chips created by a colorimetric means for outputting a color chart color signal, the density signal (Cj, Mj, Yj) of three primary main density signals (D _R j , D _G j, D _B j), and a color difference between the output of the color measurement unit and the output of the inverse masking calculation unit.
It is determined whether the color difference is the minimum, and the inverse masking coefficient is updated according to the determination result, and the control means that outputs the inverse masking coefficient that minimizes the color difference and the inverse masking coefficient output by the control means. A masking coefficient determination device comprising: an inverse function calculating means for obtaining a function and calculating a masking coefficient, and determining a masking coefficient by a convergence calculation. 2. A tristimulus color signal of color chart (X _I j, Y)
_I j, Z _I j) and outputs the output (D _R j, D _G j, D _B j) of the inverse masking calculation means to the three primary color luminance signals ( _R j, _G j, _B ).
j)) and a color conversion means for converting the three primary color luminance signals (Rj, Gj, Bj) into tristimulus color signals (X _O j, Y _O j, Z _O j). The color difference calculated by the control means is the color chart color signals (X _I j, Y _I j, Z _I j) and the color signal (X _I j
_O j, Y _O j, Z _O j). 3. The color difference calculated by the control means is L * for each of the color chart color signals (X _I j, Y _I j, Z _I j) and the color signals (X _O j, Y _O j, Z _O j). L * u * converted to coordinates in u * v * uniform color space
Euclidean distance in v * system uniform color space, or color chart color signals (X _I j, Y _I j, Z _I j) and color signals (X _O j, Y
_{3. O} j, Z _O j) are each converted into coordinates in the L * a * b * system uniform color space to obtain the Euclidean distance in the L * a * b * system uniform color space. The masking coefficient determination device described.