JP4779728B2 - How to create a color estimation model - Google Patents

Description

  The present invention relates to a method for creating a color estimation model for estimating an output value for an input value of a printer when a printer capable of recording with five or more colors including basic colors and special colors is used.

  Conventionally, color matching is known as a technique for reducing the difference in color between an image displayed on a monitor and an image recorded by a printer. In this color matching technology, there are differences in coloring depending on the printer model, so the color chart is output in advance by the printer to be recorded, each patch is measured, and output to the printer input value based on the color measurement result The value is estimated (see, for example, Patent Document 1).

In the technique of Patent Document 1, three colors of input signals such as C, M, Y, R, G, B, etc. are used, C, M, Y, K basic colors and R, G, B special colors (so-called chromatic colors). ) And a signal output in a total of seven colors. In particular, in the technique of Patent Document 1, four colors are selected from a total of seven colors to form a plurality of divided color gamuts, and these are combined to try to reproduce colors between the input value and output value of the printer. Yes.
JP 2001-136401 A

The technique of selecting a specific color from a plurality of colors and forming a plurality of divided color gamuts as in the technique of Patent Document 1 is effective in reducing the number of measurement points of patches in a color chart. For example, when using a printer capable of recording six colors Y, M, C, K, V, and O, Y, M, C, K and C, M, V, K, and Y, M, O, Color measurement for each combination with K is required. If a color chart in which one color is divided into five is output, color measurement of about 5 ⁴ × 3 = 1845 points is required, and a further reduction in the number of measurement points is desired. In addition, when a complementary color is added to a dark part of a color to be reproduced to configure an output color, the complementary color is not output as a patch in the color chart, so it is difficult to estimate an output value with respect to an input value of the printer.

  An object of the present invention is to estimate an output value with respect to an input value of a printer while keeping the number of measurement points (number of patches) in a color chart low when using a printer capable of recording with five or more colors including special colors. Is to improve the accuracy of prediction.

In order to solve the above problem, the invention according to claim 1
A method for creating a color estimation model for estimating an output value for an input value of a printer when using a printer capable of recording with five or more colors including a basic color and a special color other than the basic color,
A first output step of outputting a first color chart;
After the first output step, a first color measurement step of measuring the first color chart to obtain a first color measurement value;
A second output step of outputting a second color chart having a smaller number of colors than that of the first color chart and a larger number of gradation divisions for each color;
After the second output step, a second colorimetric step of measuring the second color chart to obtain a second colorimetric value;
An estimation step of estimating an output value with respect to an input value of the printer based on the first color measurement value and the second color measurement value after the first and second color measurement steps;
Bei to give a,
As the first selection color when outputting the second color chart, two colors having the largest difference in hue angle between adjacent primary colors among the primary colors of the basic color and the primary color of the special color are selected. A first selection step of selecting a color to be included is provided .

The invention described in claim 2
The method for creating a color estimation model according to claim 1 ,
The second output step includes
A second selection step of selecting a predetermined number of colors having the most uniform hue angles in relation to the first selection color as the second selection color when outputting the second color chart; It is characterized by providing.

The invention according to claim 3
The method for creating a color estimation model according to claim 1 ,
The second output step includes
A color distribution diagram is displayed for the combination of the first selected color and the other colors, and a third selected color is selected based on the color distribution diagram when outputting the second color chart. It is characterized by comprising a selection step.

The invention according to claim 4
In the creation method of the color estimation model according to claim 1 or 2 ,
The hue angle is defined in a uniform color space.

The invention described in claim 5
In the preparation method of the color estimation model as described in any one of Claims 1-4 ,
In the first output step, a limitation is imposed on the output value when the first color chart is output, and the first color chart is configured only by patches that satisfy the limitation. .

The invention described in claim 6
In the creation method of the color estimation model according to any one of claims 1 to 5 ,
In the second output step, a limitation is imposed on the output value when the second color chart is output, and the second color chart is configured only by patches that satisfy the limitation. .

  In the first aspect of the invention, the output value relative to the input value of the printer is estimated while keeping the number of measurement points (number of patches) in the color chart obtained by adding the first color chart and the second color chart low. The prediction accuracy when doing this can be improved.

According to the second aspect of the present invention, it is possible to efficiently secure a range of colors expressed by three or more chromatic colors, and to efficiently improve the prediction accuracy of the output value with respect to the input value of the printer with respect to the number of measurement points. Can do.

According to the third aspect of the present invention, it is possible to effectively improve the prediction accuracy of the output value of the color expression portion which is emphasized by the user.

According to the fourth aspect of the invention, the correspondence with vision is good, and the prediction accuracy when estimating the output value when viewed as a color space can be improved.

According to the fifth aspect of the present invention, the number of patches that do not substantially contribute to the estimation of the output value in the first color chart can be reduced, and the number of steps used for recording / colorimetry of the first color chart is reduced. can do.

In the invention according to claim 6 , the number of patches that do not substantially contribute to the estimation of the output value in the second color chart can be reduced, and the number of steps used for recording and colorimetry of the second color chart is reduced. can do.

  The best mode for carrying out the present invention will be described below with reference to the drawings. However, although various technically preferable limitations for carrying out the present invention are given to the embodiments described below, the scope of the invention is not limited to the following embodiments and illustrated examples.

[First Embodiment]
A method for creating a color estimation model according to the present invention is a method for estimating an output value with respect to an input value of a printer when a printer capable of recording with five or more colors including a basic color and a special color other than the basic color is used. It is.
The method mainly includes (1) a first output process for outputting a first color chart, and (2) a first color chart by measuring the first color chart after the first output process. A first color measurement step for obtaining a color value; and (3) a second output step for outputting a second color chart having a smaller number of colors and a larger number of gradation divisions for each color than the first color chart; (4) after the second output step, a second color measurement step of measuring the second color chart to obtain a second colorimetric value; and (5) first and second colorimetry steps. After that, the process is composed of five steps: an estimation step for estimating an output value for an input value of the printer based on the first color measurement value and the second color measurement value.

Hereinafter, each process of said (1)-(5) is demonstrated for every process, respectively.
In the first embodiment, Y (yellow), M (magenta), C (cyan), and K (black) are applied as basic colors, and R (red) and O (orange) are applied as special colors. The case where these six colors are input colors will be described as an example.

(1) First output step In the first and second output steps, first, the gradation of a single color is adjusted in each of the colors Y, M, C, K, R, and O (tone adjustment in the same hue). After that, the number of measurement points necessary to form grid points by dividing the color axes of the colors Y, M, C, K, R, and O is considered.

If the input of each color is 8 bits (= 2 ⁸ = 256), the division interval Δ of each color is 32, and the division number of the color axis of each color is 9, six colors of Y, M, C, K, R, and O are obtained. When combined, it is necessary to measure the color chart having 9 ⁶ = 531441 patches. This is actually difficult to measure, and even if measurement is possible, the color chart with such a large number of measurement points (number of patches) has a wide recording area, and errors due to differences in recording positions increase. If the color chart is recorded over several sheets of paper, the output value relative to the input value of the printer is not necessarily required even if the number of patches in the color chart increases due to variations (errors) in each recorded product due to repeated recording. It cannot be said that the prediction accuracy will improve.

Here, when considering a combination that equally divides the color axes of the colors Y, M, C, K, R, and O, if the number of divisions is 3, the number of measurement points is 3 ⁶ = 729, and the number of divisions Is 5, the number of measurement points is 5 ⁶ = 15625, and when the number of divisions is about 3 or 5, the number of measurement points is generally within a measurable range.

In reality, the number of measurement points is preferably within 1500 points, and more preferably within 1000 points.
In the first output step, it is determined that a first color chart having 3 ⁶ = 729 measurement points, where the number of divisions of the color axis of each color is 3, is created.

When the number of measurement points in the color chart 3 ⁶ = 729 points and determined, Y, M, C, K , R, combines a gradation value of each color of O triplicate (0,128,255) 3 ⁶ = 729 A “first color chart” having point patches is output and recorded by an ink jet printer.

  In this case, when the gradation value is 128 so that the lightness of each color is proportionally increased or decreased with respect to the input value, the ink of the gradation value is set to have an intermediate lightness between the ground color and the solid color. Record after correcting the gradation.

For patches whose total amount of ink exceeds the allowable amount (for example, 300%), K ink is retained, and Y, M, C, R, and O inks exclude the K ink amount from the allowable amount. The amount is reduced proportionally so that the total amount becomes the total amount. That is, the ink amount Y ′ of Y is based on the following formula (1).
Y ′ = (255 × 3-K) * Y / (Y + M + C + O + R) (1)
The ink amounts M ′, C ′, R ′, and O ′ of M, C, R, and O other than the Y ink amount Y ′ are also based on the above formula (1).

  In the first output step, patches are not recorded in the first color chart for which the total amount of ink exceeds the allowable amount, and the colorimetric values of the patches are measured for the patches whose input color combinations are close. You may estimate from a color value. In this case, when there are a plurality of patches that are estimation candidates, the colorimetric values of the patches that are not recorded may be the average value of the colorimetric values of the plurality of patches (see FIG. 1, in which each color is divided into five). The case where the allowable amount of ink is 250% is described.)

  In this way, when the first color chart is output / recorded, if the first color chart is configured only by patches that satisfy the restriction by limiting the output value, the first color chart to be recorded / measured The number of patches that do not substantially contribute to the estimation of the output value with respect to the input value of the printer in one color chart can be reduced, and the number of steps for recording and colorimetry of the first color chart can be reduced. .

(2) First Color Measurement Step In the first color measurement step, each patch in the first color chart is measured to obtain a “first color measurement value”.

(3) Second Output Step In the second output step, it is considered to increase the number of patches of the color chart to be measured and improve the prediction accuracy when estimating the output value for the printer input value.

  First, simply increasing the number of divisions for each color axis of Y, M, C, K, R, and O significantly increases the number of measurement points of the color chart to be measured, so that Y, M, C, K, Consider reducing the number of colors among the six colors R and O.

FIG. 2 is a distribution diagram in which the first colorimetric value is converted into a CIEL ^* a ^* b ^* color expression method value and expressed in L ^* (lightness) and a ^* (chromaticity).
As shown in FIG. 2, since the conversion values are slightly scattered at positions with high L ^* (brightness) and tend to be biased at low positions, Y, M, C, K, R, K is excluded from the six colors of O.

  Next, consider increasing the number of divisions of the color axis for each of the remaining colors Y, M, C, R, and O other than K.

In the second output step, the number of divisions for each color axis is set to 5, and the total number of measurement points including the number of newly added measurement points is set to 1000 or less.
In this case, if the colors to be newly added are five colors of Y, M, C, R, and O, the number of measurement points to be added is 5 ⁵ = 3125 points, and in this case, the newly added grid points Just over 1000 points already. Even when the color to be newly added is any one of Y, M, C, R, and O, the number of measurement points to be added is 5 ⁴ = 625 points, and the initial 3 ⁶ = 729 points is added. The total number of measurement points included exceeds 1000.

On the other hand, when the color to be newly added is any one of Y, M, C, R, and O, the number of measurement points to be added is 5 ³ = 125, and the initial 3 ⁶ = 729. The total number of measurement points including the points is 854 (= 729 + 125), which is within 1000 points. Therefore, a combination of three colors of Y, M, C, R, and O is selected as a combination that realizes the setting of the total number of measurement points within 1000 points, and the prediction accuracy of the output value is selected with the combination of the three colors. Consider improvement.

When three colors are selected from the five colors Y, M, C, R, and O, there are ₅ C ₃ = 10 options, specifically, (Y, M, C), (Y, M , R), (Y, M, O), (Y, C, R), (Y, C, O), (Y, R, O), (M, C, R), (M, C, O) ), (C, R, O), and (M, R, O).

  In the second output process, when three colors are selected from the five colors Y, M, C, R, and O, the colors to be selected are divided into the following two processes, the first and second selection processes. Each is to be selected.

(3.1) First Selection Step In the first selection step, first, by using an ink jet printer, the primary colors when each color material of Y, M, C, R, and O is used 100% are set for each color. Recording and colorimetry are performed, and each colorimetric value is converted into a color expression method value of CIEL ^* a ^* b ^* to calculate a hue angle (°) for each color. The calculation results are shown in Table 1 below.

In Table 1, “Hue” indicating the hue angle is calculated according to the following equation (2) as defined in the uniform color space.
Hue = arctan (b ^* / a ^* ) (2)

  When calculating the hue angle, if the uniform color space is adopted as in the above equation (2), the correspondence with the vision is improved, and the prediction accuracy when estimating the output value when viewed as the color space can be improved. it can.

  When the hue angles of the five colors Y, M, C, R, and O are calculated, a combination of two colors in which the hue angles of the adjacent colors between these five colors are the farthest from each other is selected. The difference in hue angle between adjacent colors among the five colors Y, M, C, R, and O is shown in Table 2 below.

  As shown in Table 2, in the difference in hue angle between the five colors Y, M, C, R, and O, the combination of (Y, C) is the largest at 161 °, and the above 10 options are available. (Y, C) is selected and determined as the first selection color included therein. As a result, ten options can be narrowed down to three options (Y, M, C), (Y, C, R), and (Y, C, O).

(3.2) Second Selection Step In the second selection step, the difference between the hue angles of the other colors (M, R, O) with respect to the two colors Y and C in the choices narrowed down in three ways. Is calculated for each color based on the hue angle in Table 1 above, and one color (any one of M, R, and O) having the same hue angle in relation to the two colors Y and C is obtained. Select / determine as the second selection color. Table 3 below shows differences in hue angles of M, R, and O for the two colors Y and C.

  As shown in Table 3, the difference between the hue angles of M, R, and O with respect to the two colors Y and C is that R is 35 °, the difference being the smallest, and the second selection for the combination of (Y, C) Select and determine R as the color. As a result, the above 10 options can be finally narrowed down to (Y, C, R) options.

When the combination of the three colors to be newly added in the five colors Y, M, C, R, and O is determined as (Y, C, R) by the above processing of the first and second selection steps, Output “second color chart” with 5 ³ = 125 patches, combining 5 gradation values (0, 64, 128, 192, 255) for each color of Y, C, and R with an inkjet printer. Record.

  In this case, as in the first output step, recording is performed after correcting the gradation of each color, and an allowable amount is set for the total amount of ink, and the ink amount of each color is based on the above equation (1). To reduce proportionally. Of course, even in the second output step, the patches in which the total amount of ink exceeds the permissible amount in the second color chart are not recorded, and the colorimetric values of the patches are measured for the patches whose input color combinations are close. You may estimate from a color value. At this time, when there are a plurality of patches that are estimation candidates, the colorimetric values of the patches that are not recorded may be the average value of the colorimetric values of the plurality of patches (see FIG. 1).

  As described above, even when the second color chart is output / recorded, if the second color chart is configured only by patches that satisfy the restriction by limiting the output value, recording / colorimetry should be performed. In the second color chart, the number of patches that do not substantially contribute to the estimation of the output value with respect to the input value of the printer can be reduced, and the number of steps used for recording and colorimetry of the second color chart can be reduced. it can.

  In the second output step, if the allowable amount for the total amount of ink is set to 300% as illustrated in the first output step, the total amount of ink always satisfies the allowable amount and the allowable amount is set. However, in the first and second output processes, the allowable amount set for the total amount of ink may be appropriately changed for each process.

(4) Second Color Measurement Step In the second color measurement step, each patch in the second color chart is measured to obtain a “second color measurement value”.

(5) Estimation Step In the estimation step, an output value for the printer input value is estimated in either the following first estimation step or the following second estimation step.

(5.1) First Estimation Step In the first estimation step, first, 3 ⁶ = 729 first colorimetric values and 5 ³ = 125 second colorimetric values are used as grid points. The grid points themselves are expanded (interpolated) by a known interpolation method (see Japanese Patent No. 2554366, etc.) to create a lookup table.

For example, 3 of ⁶ = 729 points in the first colorimetric value, 3 ⁶ = 729 → 5 ⁶ = 15625 → 9 ⁶ = and so 531441, Y, M, C, K, R, 6 colors O of A lookup table is created in which the number of grid points on each color axis is expanded from 3 points → 5 points → 9 points.

Here, the addition of 5 ³ = 125 measurement points is performed by expanding the grid points of 3 ⁶ = 729 points to 5 ⁶ = 15625 grid points, and then measuring the values of the corresponding 125 grid points. This is done by replacing Thereafter, the replaced / updated grid points of 5 ⁶ = 15625 are expanded to 9 ⁶ = 531441 grid points.

  As an explanation of extending the number of grid points on each of the six color axes Y, M, C, K, R, and O, a case of extending the grid points from 3 to 5 will be described.

Now, for LUT array LUTX [Y] [M] [C] [K] [R] [O], Y, M, C, K, R, and O are integers of 0 to 4, and each color has two subscripts. When classified by e = {0, 2, 4} and o = {1, 3}, the combination of 3 ⁶ = 729 existing as measurement points is LUTX [Ye] [Me] [Ce] [Ke] [Re ] [Oe]. The remaining points to be estimated are lattice points in which at least one of the subscripts for each color is replaced from e to o.
These points are estimated by the following processes (I) to (VI).

(I) Interpolation estimation is performed with respect to a grid point where any one color is o.
For example, in the case of a grid point of o with respect to Y, Lagrange interpolation is performed in the Y-axis direction from a combination of Ye, Me, Ce, Ke, Re, Oe = 0, 2, 4 for each of Yo = 1, 3. Estimate the value of LUTX [Yo] [Me] [Ce] [Ke] [Re] [Oe] (i). When estimating the interpolation values of M, C, K, R, and O, the following values (ii) to (vi) are estimated in the same manner.
LUTX [Ye] [Mo] [Ce] [Ke] [Re] [Oe] (ii)
LUTX [Ye] [Me] [Co] [Ke] [Re] [Oe] (iii)
LUTX [Ye] [Me] [Ce] [Ko] [Re] [Oe] (iv)
LUTX [Ye] [Me] [Ce] [Ke] [Ro] [Oe] (v)
LUTX [Ye] [Me] [Ce] [Ke] [Re] [Oo] (vi)

(II) Interpolation estimation is performed with respect to a grid point where any two colors are o.
In this case, it estimates using the result of said (I). For example, in the case of o grid points for Y and M, the result of Lagrangian interpolation in the M direction using the value (i) above and the result of Lagrange interpolation in the Y direction using the value (ii) above are 2 The estimated values from the direction are averaged to estimate LUTX [Yo] [Mo] [Ce] [Ke] [Re] [Oe]. Interpolation estimation is also performed in the same manner with respect to grid points where any two colors other than Yo and Mo are o.

(III) Similarly, using the results of (I) and (II), interpolation estimation is performed from three directions with respect to a lattice point where any three colors are o.
(IV) Similarly, using the results of (I) to (III), interpolation estimation is performed from four directions with respect to a lattice point where any four colors are o.
(V) Similarly, using the results of (I) to (IV), interpolation estimation is performed from five directions with respect to a lattice point where any five colors are o.
(VI) Similarly, using the results of (I) to (V), interpolation estimation is performed from six directions on a grid point of six colors o.

As described above, the number of grid points on each of the six color axes Y, M, C, K, R, and O can be expanded from 3 points to 5 points.
The grid points corresponding to the 5 ³ = 125 colorimetric points measured additionally are replaced at this stage.

  After that, Y, M, C, K, R, and O are set to integers of 0 to 8 as in the previous processing, and each color has two types of subscripts e = {0, 2, 4, 6, 8}, o = { 1, 3, 5, 7}, the number of grid points in each of the six color axes Y, M, C, K, R, and O is changed from 5 to 9 And can be expanded.

By executing the above processing, it is possible to create a lookup table of output values related to grid points corresponding to the input values of 9 ⁶ = 531441.
Although the output value is represented in one dimension as LUTX, since the colorimetric value is three-dimensional, LUTY and LUTZ are estimated in the same manner.

(5.2) Second Estimation Step In the second estimation step, the following interpolation is performed for input values of arbitrary six colors from the 9 ⁶ = 531441 lattice points finally obtained in the first estimation step. The output value with respect to the input value of the printer can be estimated by this method.

The LUT obtained in the above (5.1) is again expressed as LUTX [Y] [M] [C] [K] [R] [O] (vii).
In the array (vii), Y, M, C, K, R, and O indicate the grid point number of each color, and the value of LUTX indicates the value of the grid point X indicated by the grid point number.
At this time, the output value (X value) for an arbitrary input can be estimated according to the following equation (3). In the following formula (3), wi (i = Y, M, C, K, R, O) is a weighting coefficient from the lattice point, and Δ is the lattice point interval.

In the above formula (3), when δi = 0, wi = Δ− (i% Δ), and when δi = 1, wi = i% Δ. In the above formula (3), Yq = Y / Δ, Mq = M / Δ, Cq = C / Δ, Kq = K / Δ, Rq = R / Δ, and Oq = O / Δ.
Note that Y and Z values may be estimated as output values in the same manner as the X value is estimated.

  According to the first embodiment described above, since the output value for the printer input value is estimated based on the first colorimetric value and the second colorimetric value in the estimation step, the first color chart and the first color chart It is possible to improve the prediction accuracy when estimating the output value for the input value of the printer while keeping the number of measurement points (number of patches) of the color chart obtained by adding the two color charts low.

This is a case where the CMYKRO color separation values for the input grid points 9 ³ = 729 points obtained by dividing the R (red), G (green), and B (blue) colors into nine equal parts are calculated using a known RGB-CMYKRO conversion model. I will explain it.
FIG. 3 is a schematic diagram for verifying the prediction accuracy when the output value with respect to the input value of the printer is estimated. Specifically, the color separation value calculated by the RGB-CMYKRO conversion model that can be arbitrarily set. 5 is a diagram schematically representing processing until each result is compared between when the output value actually recorded by the ink jet printer is measured and when the output value is estimated by the processing of the estimation step.

As shown in FIG. 3, (i) a result (L1 ^* , a1) in which the color separation value is output and measured by an inkjet printer, and the colorimetric value is converted into a value of the color expression method of CIEL ^* a ^* b ^{*. *} , B1 ^* ) and (ii) a result (L2 ^* , a2 ^* , b2 ^* ) obtained by estimating the output value in the above estimation process using only the first colorimetric value with the color separation value as an input value. In comparison, as shown in the upper diagram of FIG. 4, the estimated error ΔE ^* (= ((L1 ^* −L2 ^* ) ² + (a1 ^* −a2) in the color region (see circles) corresponding to the specific patch numbers 100 to 200). ^* ) ² + (b1 ^* −b2 ^* ) ² ) ^1/2 ) is large. When this is plotted on the chromaticity diagram, it can be seen that the estimation error ΔE ^* is large in the color region near G as shown in the lower diagram of FIG.

On the other hand, (i) a result (L1 ^* , a1 ^* , b1) in which the color separation value is output / colorimetrically measured by an ink jet printer and the colorimetric value is converted into a value of the color expression method of CIEL ^* a ^* b ^{*. *} ) And (iii) the result of estimating the output value (L3 ^* , a3 ^{*) in the} estimation step using both the first colorimetric value and the second colorimetric value using the color separation value as an input value ^. , comparing b3 ^*) and, as in the right diagram of FIG. 5, the estimation of the color region corresponding to a particular patch number 100-200 error ^{ΔE * (= ((L1 *} -L3 *) 2 + (a1 * ^{-a3 *) 2 + (b1 *} -b3 *) 2) 1/2) is smaller.

  As described above, according to the first embodiment, the prediction accuracy when estimating the output value with respect to the input value of the printer is improved while keeping the number of measurement points (number of patches) in the first color chart low. I understand.

  Further, in the first selection step, a color including two colors (Y, M) having the largest difference in hue angle is selected as the first selection color, so that the prediction when estimating the output value with respect to the input value of the printer is performed. The accuracy can be further improved.

This will be described with reference to FIGS.
FIGS. 6 to 15 show that for each of the above 10 options, 5 ³ = 125 colorimetric values are converted into values of the color expression method of CIEL ^* a ^* b ^* to obtain a ^* (chromaticity) and b ^*. It is a distribution diagram expressed as (chromaticity). However, in FIG. 6 to FIG. 15, the diamond-shaped black dot is converted to the value of the color expression method of CIEL ^* a ^* b ^* by converting the first colorimetric value of 3 ⁶ = 729 points to a ^* (chromaticity) A plot expressed as b ^* (chromaticity) is shown, and a square-shaped white dot indicates a plot corresponding to 5 ³ = 125 colorimetric values newly added.

When comparing FIGS. 6 to 15, (Y, M, C), (Y, C, R), and (Y, C, O) including the two colors (Y, C) having the largest difference in hue angle are compared. In this combination, the number of measurement points added in the color region near G where the estimation error ΔE ^* is large effectively increases, and the prediction accuracy when estimating the output value relative to the input value of the printer is further improved. I understand that.

  In addition, during the second selection step, when the second second selected color following Y and C is selected, the difference in hue angle of M, R, and O with respect to these two colors Y and C is the largest. Since a small R is selected, it is possible to realize three colors that make the hues most uniform among the five colors Y, M, C, R, and O. Therefore, it is possible to efficiently secure a range of colors expressed by three or more chromatic colors, and it is possible to efficiently improve the prediction accuracy of the output value with respect to the input value of the printer with respect to the number of measurement points.

[Second Embodiment]
The second embodiment is different from the first embodiment in the following points, and has the same configuration as that of the first embodiment except for the above.
As a first condition, R2 (second red) is used instead of O as the input color.
As a second condition, the total number of measurement points in the first and second color charts is set within 1500 points.

From the second condition, in the second output step, the total number of measurement points including the number of measurement points to be newly added can be set within 1500 points.
In this case, since the colors to be newly added are five colors Y, M, C, R, and R2, the number of measurement points to be added is 5 ⁵ = 3125 points, and this is newly added in the same manner as described above. The number of grid points to be already exceeded 1000 points. When the color to be newly added is any one of Y, M, C, R, and R2, the number of measurement points to be added is 5 ⁴ = 625 points, and the initial 3 ⁶ = 729 points is added. The total number of measurement points including 1354 (= 729 + 625) points is within 1500 points.

  Therefore, in the second output step, a combination of four colors Y, M, C, R, and R2 is selected as a combination that realizes the setting that the total number of measurement points is 1500 or less, and the combination of the four colors To improve the prediction accuracy of the output value.

Using an ink-jet printer, the primary colors when 100% of each color material of Y, M, C, R, and R2 are used are recorded and measured for each color, and each colorimetric value is expressed as CIEL ^* a ^* b ^* . When the hue angle (°) is calculated for each color by converting it to the value of the system, it becomes as shown in Table 4 below, and the difference in hue angle between adjacent colors among the five colors Y, M, C, R, and R2 is As shown in Table 5 below.

  As shown in Table 5, in the difference in hue angle between the five colors Y, M, C, R, and R2, the combination of (Y, C) is 161 ° and the difference is the largest, as described above. In the first selection step, (Y, C) may be selected and determined as the first selection color.

  When the first selection color is determined, in the second selection step, the remaining two colors are selected and determined as the second selection color. In the selection, the four selected colors are selected so as to obtain a uniform hue angle as a whole. When the hue angle adjacent to the combination of colors is calculated, the following Table 6 is obtained.

  As shown in Table 6, it is (M, R2) that the hue angles of the four colors are arranged most evenly (= the smallest variation in the hue angles of the three colors) in the combination with Y and C. In this selection step, M and R2 may be selected and determined as the second selection color.

  By using the color estimation model creation method according to the first and second embodiments described above, input (for example, three colors of R, G, and B) and output (for example, Y, M, C, K, R, O, or It is possible to predict the color when assuming a color separation model between the six colors of R2. For example, based on the predicted output value, the arbitrarily settable RGB-YMCKRO conversion model or the like can be sequentially corrected and preferably improved. Alternatively, the color matching for a specific target can be performed by obtaining the relationship between the input of the conversion model and the estimated color.

  The present invention is not limited to the first and second embodiments described above, and various improvements and design changes may be made without departing from the spirit of the present invention.

  As one improvement / design change matter, instead of the second selection step, it is obtained from a color distribution chart obtained from the first colorimetric value (see FIG. 2) or a colorimetric value of a candidate color as a selected color. All the color distribution charts (see FIGS. 6 to 15) to be displayed are displayed in advance for each combination (may be displayed on a monitor or recorded on paper by a printer). A third selection step for arbitrarily selecting and determining a third selection color instead of the two selection colors may be provided. In this case, it is possible to effectively improve the prediction accuracy of the output value of the color expression portion that is important to the user.

  As other improvements / design changes, other colors may be applied instead of or in addition to Y, M, C, and K as basic colors, and other colors in place of or in addition to R and O as special colors The input colors are not limited to the six colors Y, M, C, K, R, O or R2, and any one of these colors may be used. It is good also as a color and adding seven or more colors.

  As another improvement / design change item, the input target of the input value is not limited to the ink jet printer, but may be an electrophotographic printer or a sublimation printer.

6 is a diagram illustrating a method for estimating a colorimetric value of a patch that exceeds the allowable amount from a colorimetric value of a patch that satisfies the allowable amount when the allowable amount of ink is limited. It is the distribution map which converted the 1st colorimetric value into the value of the color expression system of CIEL ^* a ^* b ^* and expressed with L ^* (lightness) and a ^* (chromaticity). It is a schematic diagram for verifying the prediction accuracy when the output value with respect to the input value of the printer is estimated. (I) Output / colorimetry of a predetermined color separation value with an ink jet printer and conversion of the colorimetry value into a value of the color expression method of CIEL ^* a ^* b ^* , and (ii) input the color separation value It is drawing explaining the comparison result with the result of having estimated the output value by the estimation process using only the 1st colorimetric value as a value. (I) Output / colorimetry of a predetermined color separation value by an ink jet printer and conversion of the colorimetry value into a value of the color expression method of CIEL ^* a ^* b ^* , and (iii) input the color separation value It is drawing explaining the comparison result with the result of estimating the output value in the estimation process using both the first colorimetric value and the second colorimetric value as values. FIG. 6 is a distribution diagram in which 125 colorimetric values of Y, M, and C are converted into values of the color expression method of CIEL ^* a ^* b ^* and expressed as a ^* (chromaticity) and b ^* (chromaticity). . FIG. 5 is a distribution diagram in which 125 colorimetric values of Y, M, and R are converted into values of the color expression method of CIEL ^* a ^* b ^* and expressed by a ^* (chromaticity) and b ^* (chromaticity). . FIG. 7 is a distribution diagram in which 125 colorimetric values of Y, M, and O are converted into values of the color expression method of CIEL ^* a ^* b ^* and expressed as a ^* (chromaticity) and b ^* (chromaticity). . FIG. 5 is a distribution diagram in which 125 colorimetric values of Y, C, and R are converted into values of a color expression method of CIEL ^* a ^* b ^* and expressed by a ^* (chromaticity) and b ^* (chromaticity). . FIG. 6 is a distribution diagram in which 125 colorimetric values of Y, C, and O are converted into values of a color expression method of CIEL ^* a ^* b ^* and expressed by a ^* (chromaticity) and b ^* (chromaticity). . FIG. 7 is a distribution diagram in which 125 colorimetric values of Y, R, and O are converted into values of the color expression method of CIEL ^* a ^* b ^* and expressed as a ^* (chromaticity) and b ^* (chromaticity). . FIG. 7 is a distribution diagram in which 125 colorimetric values of M, C, and R are converted into values of a color expression scheme of CIEL ^* a ^* b ^* and expressed as a ^* (chromaticity) and b ^* (chromaticity). . FIG. 5 is a distribution diagram in which 125 colorimetric values of M, C, and O are converted into values of the color expression method of CIEL ^* a ^* b ^* and expressed as a ^* (chromaticity) and b ^* (chromaticity). . FIG. 7 is a distribution diagram in which 125 colorimetric values of C, R, and O are converted into values of the color expression method of CIEL ^* a ^* b ^* and expressed by a ^* (chromaticity) and b ^* (chromaticity). . FIG. 5 is a distribution diagram in which 125 colorimetric values of M, R, and O are converted into values of the color expression method of CIEL ^* a ^* b ^* and expressed as a ^* (chromaticity) and b ^* (chromaticity). .

Claims (6)

A method for creating a color estimation model for estimating an output value for an input value of a printer when using a printer capable of recording with five or more colors including a basic color and a special color other than the basic color,
A first output step of outputting a first color chart;
After the first output step, a first color measurement step of measuring the first color chart to obtain a first color measurement value;
A second output step of outputting a second color chart having a smaller number of colors than that of the first color chart and a larger number of gradation divisions for each color;
After the second output step, a second colorimetric step of measuring the second color chart to obtain a second colorimetric value;
An estimation step of estimating an output value with respect to an input value of the printer based on the first color measurement value and the second color measurement value after the first and second color measurement steps;
Bei to give a,
The second output step includes
As the first selection color when outputting the second color chart, two colors having the largest difference in hue angle between adjacent primary colors among the primary colors of the basic color and the primary color of the special color are selected. A color estimation model creation method comprising a first selection step of selecting a color to be included. The method for creating a color estimation model according to claim 1 ,
The second output step includes
A second selection step of selecting a predetermined number of colors having the most uniform hue angles in relation to the first selection color as the second selection color when outputting the second color chart; A method for creating a color estimation model comprising: The method for creating a color estimation model according to claim 1 ,
The second output step includes
A color distribution diagram is displayed for the combination of the first selected color and the other colors, and a third selected color is selected based on the color distribution diagram when outputting the second color chart. A method for creating a color estimation model, comprising: In the creation method of the color estimation model according to claim 1 or 2 ,
A method for creating a color estimation model, wherein the hue angle is defined in a uniform color space. In the preparation method of the color estimation model as described in any one of Claims 1-4 ,
In the first output step, a restriction is imposed on the output value when the first color chart is output, and the first color chart is configured by only patches satisfying the restriction. How to create a color estimation model. In the creation method of the color estimation model according to any one of claims 1 to 5 ,
In the second output step, a limit is imposed on the output value when the second color chart is output, and the second color chart is configured by only patches that satisfy the limit. How to create a color estimation model.