JP2002281328A - Method for creating color conversion table - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method for creating a color conversion table with high color reproducibility. SOLUTION: A color of color patch formed by setting a lattice points of 9×9×9×9 in a CMYK color space and making a printer output it is measured, a color measuring table of CMYK-XYZ is created by calculating an XYZ value, the number of pieces of data is further increased to 25×25×25×25 by an eight- point interpolation method, a specified RGB value of 17×17×17 is converted into the XYZ value, the XYZ value is converted into an L*a*b* value, a K value in the CMYK space equivalent to the L*a*b* value is determined from an ICC profile, the lattice point at which color difference ΔE in an L*u*v* color space becomes minimum is searched from the color measuring table without changing the K value and a CMY value of the lattice point is determined as a value corresponding to the XYZ value.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for creating a color conversion table, and more particularly, to a method for converting a color signal once converted into a coordinate value in an intermediate three-dimensional color space which does not depend on the color reproduction characteristics of a device, from an output device. The present invention relates to a method for creating a color conversion table for converting coordinate values in a four-dimensional color space according to characteristics.

[0002]

2. Description of the Related Art When printing or displaying a color image, usually, first, an original such as a photograph is read by an image input device such as a scanner to generate image data. Then, by outputting this image data with an image output device such as a printer or a display, a color image is obtained. By the way, since the color reproduction characteristics of the input device and the output device are different, the colors of the original color image and the output or printed color image may be slightly different.

Therefore, conventionally, in reproducing an image, a color signal of image data dependent on a device is first converted into a color signal of an intermediate color space independent of a device, and then converted into a color space unique to an output device. The method has been adopted. Such color space conversion is performed using a profile in which the color reproduction characteristics of each device are described. Profiles are generally provided in the form of a look-up table.

[0004] An example of a method of creating a reference table for converting a color signal of image data that depends on a device into a color signal of an intermediate color space that does not depend on a device is disclosed in Japanese Patent Application Laid-Open Publication No. 200-200200.
No. 0-253270. According to the technology described in this publication, according to the distance between a color patch and a grid point, a grid point closer to a certain color patch is weighted to increase the influence of a color difference in the color patch, and is corrected. By correcting the color values of the initial lookup table only for the neighboring grid points within a predetermined distance from the patch, and repeating a series of correction processes,
Create a color conversion table efficiently.

Various multidimensional color conversion tables for converting color signals in an intermediate color space into the color space of an output device are commercially available as software for creating an ICC or the like. The multidimensional color conversion table measures the color of a color patch, searches for a grid point closest to the designated RGB based on the colorimetric data, and acquires a CMY value of the grid point in the CMYK color space. Then, perform UCR etc. and perform CMY
The K (black) value in the K color space is calculated.

[0006]

However, in the conventional multidimensional color conversion table, the CMY values are determined first, and the K value is determined later. For this reason, the CMY values and the CMYK colorimetric values are different, and the color of the searched lattice point is different from the reproduced color. As a result, there is a problem in that K (black) is excessively added to a color that should be originally bright, and the reproducibility of hue may be deteriorated.
In particular, accurate reproduction of light colors such as flesh color and gray was sometimes difficult.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-described problem, and has as its object to provide a method of creating a color conversion table having high color reproducibility.

[0008]

According to a first aspect of the present invention, there is provided a color conversion table creating method, comprising: (a) setting grid points in a CMYK color space; Measure the color patch output and XY
Obtaining a Z value and creating a CMYK-XYZ colorimetric table; (b) converting a designated RGB value to an XYZ value; and (c) converting the XYZ value to a K value in a CMYK color space.
A grid point having a minimum color difference from the XYZ value is searched from the colorimetric table without changing the K value determined in the steps (d) and (c), and the CMY values corresponding to the XYZ values are determined. The method includes a step of determining a value and a step of performing (e), (b) to (d) for all designated RGB values, and creating an RGB-CMYK color conversion table.

As described above, according to the color conversion table creating method of the present invention, the K value in the CMYK color space is determined first, and thereafter, without changing the K value, a grid point with the minimum color difference is searched. Then, the CMY value of the grid point is determined as a value corresponding to the XYZ value. Thus, according to the present invention, a color conversion table with high color reproducibility can be created.

When color conversion is performed using the color conversion table created by the method of the present invention, the amount of the black component in the CMYK color space becomes equal to the ratio in the ink of the original image, and accurate hue reproduction is achieved. Can be planned.
In particular, accurate reproduction of light colors such as flesh color and gray can be achieved.

By the way, due to the limitation of the storage capacity of the ROM or the like, the number of grid points of the multidimensional color conversion table is usually 9 × 9.
× 9, and at most 17 × 17 × 17. Therefore, in the conventional multidimensional color conversion table, the number of patch data is small, and linearity between grid points is not maintained. If interpolation is performed in a space where linearity is not maintained, the gradation or the like may be reversed. Such a reversal is particularly noticeable in gray gradations.

Therefore, according to the invention of claim 2,
In the step (a), an XYZ value corresponding to a new grid point between coordinate values in the set CMYK color space is obtained by interpolation in a colorimetric table and added.
As described above, if the colorimetric data of the colorimetric table is increased by interpolation after the colorimetric measurement, the linearity between the patch data can be maintained. As a result, by using the color conversion table created in this way, it is possible to suppress the occurrence of the inversion of the gradation.

According to the third aspect of the present invention,
In step (c), in determining the K value corresponding to the XYZ value, and converts the XYZ values into ^L * ^a * ^{b *} values, the K values in CMYK space corresponding to the ^L * ^a * ^{b *} values ICC There is a method to obtain from the profile. Thus, the K value can be easily obtained from the XYZ values.

By the way, when a color patch is measured and a grid point closest to the designated RGB, that is, a grid point with the smallest color difference, is searched based on the colorimetric data, the color difference is usually calculated as L ^* a ^*. The search is represented by b ^* value. However, in the color space of L ^* a ^* b ^* , the hue on the blue side is distorted. For this reason, it is difficult to accurately reproduce the color in the blue direction if the color difference is searched for by using the L ^* a ^* b ^* values. For example, the hue of blue may become purple.

Therefore, according to the invention described in claim 4,
In the step (d), the color difference is L^*u^*v^*In color space
Color difference. L^*u^*v^*Color space
Has no distortion in the blue hue. Therefore, the color difference is L
^*u^*v^*If you search by value, the hue in the blue direction is positive.
Reliable reproduction can be achieved.

According to the fifth aspect of the present invention,
In the step (e), a process of alternately repeating a process of decreasing the number of grid points of the RGB-CMYK color conversion table by interpolation and a process of increasing the number of lattice points to perform smoothing. If the color conversion table is smoothed in this way, the data in the color conversion table can be smoothly changed with respect to the data of the surrounding grid points. As a result, if the color conversion is performed using this color conversion table, a smooth gradation can be reproduced.

[0017]

Embodiments of the present invention will be described below with reference to the drawings. When creating the color conversion table of the present embodiment, first, grid points in the CMYK color space (color system) are set, that is, coordinate values are set in a grid pattern, and the grid point of each grid point in the CMYK color space is set in the image device. A color patch is output (step S1 in FIG. 2).

FIG. 1 schematically shows CMYK color patches. This color patch is cyan (C),
The coordinate values of each grid point in the four-dimensional color system of magenta (M), yellow (Y), and black (K) are set and output to a printer. Here, 9 × 9 × 9 ×
Since 9 grid points are set, this color patch is 656
It has one color.

Then, the color patches are measured with a colorimeter to obtain XYZ values, and a CMYK-XYZ colorimetric table is created (step S2 in FIG. 2). Here, D50
Perform colorimetry in an environment.

Subsequently, in order to increase the number of measured data,
According to the flow shown in FIG. 3, the number of data in the CMYK-XYZ colorimetric table is increased to 25 × 25 × 25 × 25 = 390625 by eight-point interpolation (step S in FIG. 3). As described above, if the colorimetric data of the colorimetric table is increased by interpolation, the linearity between patch data can be maintained.

Note that Loop1 (0 ≦ C ≦ 2) in FIG.
5), Loop2 (0 ≦ M ≦ 25), Loop3 (0 ≦ M)
Y ≦ 25) and Loop 4 (0 ≦ K ≦ 25) are sequentially assigned to each grid point obtained by dividing the coordinates of cyan (C), magenta (M), yellow (Y) and black (K) into 25. 8 is calculated based on the data of the eight grid points P1 to P8.

Next, referring to the flowchart of FIG.
An example of a method for creating a color conversion table using the color measurement table will be described. Here, 1 in the sRGB color space
The designated RGB values of each grid point divided into 7 × 17 × 17 are converted into CMYK values in a CMYK color space to create a color conversion table.

Next, the designated RGB values are converted into XYZ values (step S1 in FIG. 4). Here, FIG.
The conversion example from a value to an XYZ value is shown typically. In the example shown in FIG. 5, the designated RGB value is converted to an R′G′B ′ value by applying Gamma of 2.2 (step S1 in FIG. 5). This is in consideration of the monitor Gamma.

Subsequently, the sRGB conversion matrix is applied to the R'G'B 'values to convert them into XYZ values (D65) (FIG. 5).
Step S1). An example of the sRGB conversion matrix is shown below the block of the sRGB conversion matrix in FIG.

By the way, the color measurement of the color patch is D50.
Under the environment. Therefore, it is necessary to match the XYZ values with the light source at the time of color measurement. Therefore, the XYZ value (D6
5) is subjected to a Bradford matrix in consideration of the light source, and is converted into an XYZ value (D50) (step S3 in FIG. 5). An example of a Bradford matrix is shown below the Bradford matrix block in FIG.

Next, the K (black: black) value in the CMYK color space is determined from the XYZ values (step S2 in FIG. 4). Here, FIG. 6 is a flowchart illustrating an example of a method of determining the K value from the XYZ values. First, XYZ
In determining the K value corresponding to the value, the XYZ value is converted to L ^*
It is converted to an a ^* b ^* value (coordinate value in L ^* a ^* b ^* color space) (step S1 in FIG. 6). In this conversion, D50 is used as a light source.

Next, CM corresponding to the L ^* a ^* b ^* value
The K value in the YK space is obtained from the ICRO profile of EUROInk (step S2 in FIG. 6). EURO
By using Ink's ICC profile, K
The ratio of (black) is similar to that of ink printing.
As a result, it is possible to improve the reproducibility of light colors such as skin color and gray. Also, when searching for the K value of CMYK,
It is preferable to use an eight-point interpolation method.

Next, without changing the K value, the colorimetric table
Then, a grid point that minimizes the color difference from the XYZ values is searched, and
And the CMY values corresponding to the XYZ values
(Step S3 in FIG. 4). Here, FIG.
The following shows an example of a method for determining the CMY value. First, the XYZ value is L
^*u^*v ^*Value (L^*u^*v^*Color space)
Conversion (step S1 in FIG. 7). This transformation uses a light source
Is used as D50.

Next, 25 × 25 × 25 × 25 is obtained by interpolation.
From the colorimetric table that has been increased, the grid point where the K (black) value is equal and the color difference ΔE (L ^* u ^* v ^* ) is the smallest is found. That is, in determining the color difference, first,
The XYZ values (D50) (A in FIG. 7) of the colorimetric table are
^* UI ^* vI ^* (D50) (B in FIG. 7), and a color difference ΔE (D in FIG. 7) from the input L ^* u ^* v ^* (C in FIG. 7) is obtained (step S2 in FIG. 7). . Note that the color difference is ΔE =
{(LI-L) ² + (uI-u) ² + (vI-v) ² } ^1/2
It is represented by

Then, all the colorimetric tables are searched to find a grid point of XYZ (D50) that minimizes the color difference ΔE (L ^* u ^* v ^* ), and outputs a CMY value corresponding to the grid point (FIG. Step S3).

As described above, if the K value is determined first and the CMY value is determined by searching for a grid point with the minimum color difference without changing the K value, a light color such as flesh color or gray can be obtained. A color conversion table with high color reproducibility can be created.

Note that L^*u^*v^*The color space of the blue side
Since the hue is not distorted, the color difference is L ^*u^*v^*Expressed by value
Search for accurate hue in the blue direction
Can be. In addition, the brightness, the saturation, etc. are weighted and the color difference Δ
If E is taken, the output color can be adjusted.

Next, the CMYK values thus obtained are entered in the RGB-CMYK color conversion table (step S4 in FIG. 4). This operation is referred to as Loop in FIG.
The process is performed for all designated RGB values according to 1 to Loop3, and an RGB-CMYK color conversion table is created.

By the way, the RG created in this way is
If the B-CMYK color conversion table is left as it is, it may be difficult to reproduce a smooth gradation. Therefore, in the present embodiment, RGB-CMYK
Apply smoothing to the color conversion table.

In applying smoothing, 17
The number of grids (number of meshes) of × 17 × 17 is once converted to the number of grids of 16 × 16 × 16 by eight-point interpolation. Further, the 16 × 16 × 16 grid number is again converted to the original 17 × 16 × 16
The number of grids is returned to 17 × 17. By repeating this operation five times, for example, the data of the color conversion table becomes values that smoothly change with respect to the data of the surrounding grid points, and a smooth gradation can be reproduced.

In the embodiment described above, an example in which the present invention is configured under specific conditions has been described. However, the present invention can be variously modified. For example, in the above-described embodiment, an example has been described in which a specific number of grid points are set as the number of grid points of a color patch or a designated RGB value, but the present invention is not limited to this.

Further, in the above-described embodiment, an example has been described in which the K value is determined from the ICC profile of EURO Ink. However, in the present invention, the profile used when determining the K value is not limited to this. . For example, an ICC profile such as SWOP may be used.

[0038]

As described above in detail, according to the color conversion table creating method of the present invention, the K value in the CMYK color space is determined first, and then the K value is not changed.
A CMY value is determined by searching for a grid point that minimizes the color difference. Thus, a color conversion table with high color reproducibility can be created. If color conversion is performed using the color conversion table created by the method of the present invention, accurate hue reproduction can be achieved. In particular, accurate reproduction of light colors such as flesh color and gray can be achieved.

If the colorimetric data in the colorimetric table is increased by interpolation, the linearity between the patch data can be maintained. As a result, by using the color conversion table created in this way, it is possible to suppress the occurrence of the inversion of the gradation. Further, if the color difference is represented by the color difference in the L ^* u ^* v ^* color space and the search is performed, the color difference is L ^* a ^* b ^*.
It is possible to achieve more accurate reproduction of the hue in the blue direction than in the case represented by

[Brief description of the drawings]

FIG. 1 is a schematic diagram of a color patch.

FIG. 2 is a flowchart illustrating a process of creating a colorimetric table.

FIG. 3 is a flowchart illustrating a process of increasing the number of grid points in a colorimetric table.

FIG. 4 is a flowchart illustrating a method for creating a color conversion table according to the embodiment.

FIG. 5 is a diagram showing an example in which the sRGB value is set to X in step S1 of FIG.
It is a flowchart for demonstrating the process of converting into YZ value.

FIG. 6 is a flowchart illustrating a process of obtaining a K value in step S2 of FIG. 4;

FIG. 7 is a flowchart illustrating a process of obtaining a CMY value in step S3 of FIG. 4;

 ──────────────────────────────────────────────────続 き Continued on the front page F-term (reference) 2C262 AA24 AA26 AA27 AB11 AC02 BA02 BA07 BA17 BC01 BC15 BC19 FA13 5B057 BA02 CA01 CA08 CA12 CA16 CB01 CB08 CB12 CB16 CC01 CE17 CE18 CH07 5C077 LL05 LL12 LL19 MM27 PP08 PP33 PP36 PP38 PP43 PP47 PP66 PQ12 PQ23 RR19 5C079 HB01 HB03 HB05 HB08 HB09 HB11 LA28 LB02 MA04 MA10 MA11 NA03 NA29 PA05

Claims (5)

[Claims] (A) setting a grid point in a CMYK color space and measuring a color patch output to an image device to determine an XYZ value, and creating a CMYK-XYZ colorimetric table; b) converting a designated RGB value into an XYZ value; (c) determining a K value in a CMYK color space from the XYZ value; and (d) keeping the K value determined in the step (c). (E) searching for a grid point having a minimum color difference from the XYZ value from the colorimetric table and determining a CMY value corresponding to the XYZ value; The process is performed using all designated RGB
Creating a RGB-CMYK color conversion table by performing the process on the values. 2. In the step (a), an XYZ value corresponding to a new grid point between coordinate values in a set CMYK color space is obtained by interpolation and added to the colorimetric table. The method according to claim 1. 3. A step (c), in determining the K value corresponding to the XYZ value, and converts the XYZ values into ^L * ^a * ^{b *} values, the ^L * ^{a *}
3. A K value in a CMYK space corresponding to a b ^* value is obtained from an ICC profile.
The described color conversion table creation method. 4. In the step (d), the color difference is set to L
4. A method according to claim 1, wherein the color difference is a color difference in a ^* u ^* v ^* color space. 5. In the step (e), the RGB-
5. The method according to claim 1, wherein a process for decreasing the number of grid points of the CMYK color conversion table by interpolation and a process for increasing the number are alternately repeated to apply smoothing.