JP2005144974A - Device, method and program for correcting color conversion table and printer controller - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To form a color conversion table for performing color conversion with high accuracy by correcting the color conversion table easily and surely without being affected by artificial factors, e.g. skill of a correcting person. <P>SOLUTION: A color conversion table defining correspondence between color data in a predetermined color space and ink quantity data indicative of the quantity of ink of each color being used in a printer is acquired, target data indicative of the ideal value of a color being printed by predetermined ink quantity data is acquired, a patch is printed by the predetermined ink quantity data through a printer, colorimetric value is acquired by performing colorimetry on the printed patch, a correction vector is calculated based on a vector having a reverse direction component of a difference vector indicative of the difference between the ideal value and the colorimetric value in a predetermined machine independent color space, ink quantity data for reducing the difference by the correction vector is calculated, and then the predetermined ink quantity data is corrected by the ink quantity data thus calculated. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a color conversion table correction apparatus, a color conversion table correction method, a color conversion table correction program, and a print control apparatus.

In recent years, inkjet printers, laser printers, and the like have become widespread as color printers. A color printer uses a color conversion table that converts input color image data into a combination of a plurality of ink amounts. As the color conversion table, for example, a lookup table in which input color image data is associated with a combination of a plurality of ink amounts is often used. In the conventional method for creating a color conversion table, (i) coordinate values in a device-independent color space corresponding to a large number of input color image data are obtained, and (ii) multiple color patches are printed with multiple ink amount data. , (Iii) measure the color of each color patch to obtain the coordinate value in the device-independent color space, and (iv) input color image data and ink amount data from the correspondence between the coordinate values in the device-independent color space. A look-up table representing the correspondence between and was created. (For example, refer to Patent Document 1).
JP 2002-118762 A

The conventional color conversion program described above has the following problems.
That is, the actual number of operations for printing and measuring the color patches is limited, and a color conversion error occurs due to the limited number of color patches. In particular, this color conversion error cannot be ignored for a color whose color change is easy to understand. For example, the actual number of patches for which printing and colorimetry can be performed is about 10 ³ , but in the color conversion table, the correspondence relationship is defined for a much larger number of colors. Accordingly, the coordinate values of these colors are calculated by an interpolation calculation using the colorimetric values of the color patches. In such an interpolation, an error always occurs. Of course, the color conversion table is created in a state including various errors such as a colorimetric error.

Therefore, even if color conversion is performed with reference to the color conversion table and printing is performed, a color exactly equal to the color indicated by the input color image data is not always printed. For colors such as gray that are easy to recognize color changes, high quality printing cannot be performed if there is such a difference. In order to eliminate such differences, it is conceivable to artificially correct the ink amount data specified in the color conversion table. However, artificial correction requires the skill of the corrector, and the corrector must Variation will occur.
The present invention has been made in view of the above-mentioned problems, and is a color that performs high-precision color conversion by correcting the color conversion table easily and accurately without being affected by human factors such as the skill of the corrector. It is an object of the present invention to provide a color conversion table correction apparatus, a color conversion table correction method, a color conversion table correction program, and a print control apparatus that can create a conversion table.

  In order to achieve at least one of the above objects, the invention according to claim 1 corrects specific ink amount data for a color conversion table defined in advance. That is, the ink amount data is corrected so that the color data in a predetermined color space is converted into ink amount data different from the original color conversion table for a specific color. This correction is realized by a correction vector calculated based on a vector having a reverse component of a difference vector indicating a difference between the ideal value and the colorimetric value. That is, the correction is performed by calculating ink amount data for reducing the difference using the correction vector, and updating the specific ink amount data with the calculated ink amount data. Therefore, when color conversion is performed using the corrected color conversion table, as specific ink amount data, data that can be printed in a state where the difference between the ideal value and the colorimetric value is very small is obtained. For this reason, it is possible to perform printing by performing color conversion with very high accuracy.

  Here, the table acquired by the color conversion table data acquisition unit may be prepared in advance, but the acquired color conversion table data is a table in which the accuracy of color conversion is improved by the correction according to the present invention. Therefore, it is not a color conversion table data in which the ideal value and the colorimetric value are strictly equal for all ink amount data, but a table in which a certain degree of error can occur. Further, in the present invention, in order to improve the color conversion accuracy for specific ink amount data, it is preferable to apply the present invention when color conversion with high accuracy is particularly required for the color indicated by the specific ink amount data. .

  The color conversion table data only needs to be able to convert color data in a predetermined color space into ink amount data. For this purpose, the color data in the predetermined color space and the ink of each color used in the printing apparatus are used. It is only necessary to define a correspondence relationship with the ink amount data indicating the amount. Therefore, it may be data in which the color data and the ink amount data are associated with each other for a plurality of colors, or profile data that defines the relationship between both based on a plurality of parameters.

Further, the predetermined color space may be a color space of a color system different from the ink amount space including the amount of ink used in the printing device as a component, and the color space or device used in an image device different from the printing device. Various color spaces such as an independent color space can be employed. For example, an sRGB space used for a display or an RGB space used for a scanner can be employed as the former, and an L ^* a ^* b ^* space, an L ^* u ^* v ^* space, or the like can be employed as the latter. . Of course, the color data in the predetermined color space only needs to be data for specifying the color in the color space, and data representing the size of each color component in each color system can be adopted. It is.

  The target data only needs to be able to indicate an ideal value of a color to be printed with specific ink amount data. In other words, it is difficult to create a color conversion table so that the ink amount data obtained by color conversion referring to the color conversion table data is exactly the same as the color before color conversion. It is possible to determine in advance exactly whether to convert to. Therefore, in the present invention, the color to be obtained as a result of the color conversion is set as an ideal value and defined in advance as target data.

  The color printed by the specific ink amount data can be specified by printing by the patch printing unit and the colorimetric value obtained by the colorimetric value acquisition unit. Accordingly, the patch printing unit only needs to be able to control the printing apparatus so as to print a patch having a predetermined area based on the ink amount data, and the colorimetric value acquisition unit can measure the colorimetry acquired by a colorimeter or the like. It is sufficient if the color value can be acquired.

  The ink amount data correction means uses the difference between the ideal value and the colorimetric value in a predetermined device-independent color space when performing correction so that the output color based on the corrected ink amount data approaches the target. . Therefore, it is preferable that the ideal value and the colorimetric value are acquired as values in the device-independent color space. Of course, it may be configured to obtain a value in the device-independent color space through some conversion. Further, after the specific ink amount data is corrected, the exact color is not known unless the result of printing with the corrected ink amount data is measured. However, if the corrected color can be specified in a predetermined device-independent color space, the ink amount data corresponding to the corrected color and at least the difference is highly likely to be eliminated. Can be predicted.

  For example, if the correspondence between the device-independent color data in the predetermined device-independent color space and the ink amount data is defined in advance, the ink amount data corresponding to the corrected color is predicted by interpolation calculation or the like. can do. Therefore, if the corrected color in a predetermined device-independent color space is determined so that the difference between the color indicated by the corrected ink amount data and the ideal value is eliminated, the correction in the ink amount data correcting unit is performed. Can be implemented. Of course, in this correction, it is only necessary to perform correction so as to eliminate the difference between the ideal value and the colorimetric value within a predetermined device-independent color space, and various configurations can be employed.

For example, since there is a difference between the ideal value and the colorimetric value, if the correction vector has at least a component in the reverse direction with respect to the difference vector representing the difference, the ideal value and the color measurement value It can be considered that the direction is to eliminate the difference from the colorimetric value. Therefore, under such an idea, the corrected color is determined by the sum of a correction vector having at least a component in the reverse direction with respect to the difference vector from the ideal value to the colorimetric value and a vector indicating the ideal value. Correction may be performed. As the difference vector, in addition to a vector indicating the difference between the ideal value and the colorimetric value, a vector that is a scalar constant multiple can be adopted. As the predetermined device-independent color space, an L ^* a ^* b ^* space, an L ^* u ^* v ^* space, or the like can be used.

  Of course, this correction need not be completed only once. For example, after the correction by the ink amount data correction unit, if the printing by the patch printing unit and the colorimetric value acquisition by the colorimetric value acquisition unit are performed based on the corrected ink amount data, the correction is performed again. Can be implemented. That is, by repeating the correction, the difference between the colorimetric value and the ideal value can be gradually eliminated, and the ink amount data can be corrected so that the two are finally very close.

  Various methods can be adopted as a specific correction method when the correction is repeated. For example, the difference between the colorimetric value of a patch printed with the ink amount data obtained by the Nth correction (N is a natural number) and the value of the device-independent color space indicating the color (the value after the Nth correction) The correction may be performed with reference to the difference vector in the N-th correction and the difference vector in the M-th correction (M is a natural number equal to or less than N−1). That is, when only the difference between the color after the Nth correction and the colorimetric value is considered in the Nth correction, the Nth correction is very close to the ink amount data after the Mth correction. When the value is obtained, the correction after the (N + 1) th time is repeated substantially the same as the previous correction. That is, the difference may not be resolved (so-called vibration) regardless of how many times correction is performed.

  However, if the N-th correction is performed with reference to the difference in the correction stage before the N-1th correction, even if the ink amount data after the M-th correction is obtained in the N-th correction, a value very close to that is obtained. In the (N + 1) th and subsequent corrections, the difference of NM times is taken into consideration, so that substantially the same correction as the Mth correction is not repeated. Therefore, it is possible to effectively prevent the occurrence of vibration and perform correction for quickly eliminating the above-described difference. Further, when only the difference between the color after the N-th correction and the colorimetric value for the ink amount data is taken into consideration in the N-th correction, the N-th correction is caused by an interpolation error, a colorimetric error, or the like. The above-mentioned difference may be increased by so-called correction (so-called divergence), but the occurrence of such divergence can be effectively prevented by referring to the difference before the (N-1) th time.

  Furthermore, as a specific method for determining the correction vector while iterating, correction is performed by multiplying the inverse vector of the difference vector by adding a weighting factor so that the smaller the difference, the greater the contribution to correction. The difference vector may be a difference vector indicating a difference between the colorimetric value and the ideal value, or a device-independent color space value indicating a color corrected by the ink amount data correcting unit and its measurement. The difference vector from the color value). That is, in considering the difference before the (N-1) th time in the Nth correction, the inverse vector of the difference vector at each correction stage is considered, and the smaller the inverse vector size is, the larger the value is for each inverse vector. Multiply by the coefficient A vector obtained by linear combination is set as a correction vector.

  According to such determination of the correction vector, the smaller the difference in the respective correction stages, the larger the difference is considered. In the present invention, as described above, the corrected color is determined in a predetermined device-independent color space so that the difference between the colorimetric value of the color indicated by the corrected ink amount data and the ideal value is eliminated. However, in this correction, there is a higher possibility that the ink amount data can be determined more accurately as the difference between the colorimetric value of the color indicated by the corrected ink amount data and the ideal value is smaller. For example, correspondence data indicating the correspondence between the device-independent color data in the predetermined device-independent color space and the ink amount data is defined in advance, and the device-independent color data corrected by tetrahedral interpolation is used. When calculating the ink amount data, if the difference is small, there is a high possibility that the corrected color and the ideal value exist in the same tetrahedron.

  Within the same tetrahedron, the relationship between device-independent color data and ink amount data is linear, but if they are in different tetrahedra, the relationship between device-independent color data and ink amount data is nonlinear at the tetrahedron boundary. Therefore, when performing correction based on the difference, it becomes difficult to evaluate the difference, and an error is likely to occur. Therefore, if the ideal value and the colorimetric value of the corrected color are in the same tetrahedron, accurate interpolation is possible, and there is a high possibility that the difference is quickly eliminated. Of course, in terms of the accuracy of interpolation, it is not limited to tetrahedral interpolation, and in other interpolations, the smaller the difference, the higher the possibility of accuracy.

  That is, according to the linear combination based on the weights as described above, the contribution to the determination of the correction vector can be increased as the difference vector of the data is more likely to be accurately interpolated. As a result, the difference can be eliminated more quickly while preventing the occurrence of vibration and divergence as described above. Further, in determining such a correction vector, the correction vector can be determined based on the colorimetric value and the ideal value. Accordingly, there is no room to be influenced by the subjectivity and skill of the corrector, and the color conversion table can be corrected easily and accurately without being affected by human factors such as the skill of the corrector.

  The above-mentioned prediction can be carried out by previously defining the correspondence data as described above. However, the correspondence data is measured when the color conversion table is created. It is preferable that the correspondence relationship is defined for a larger number of colors than the number of colors. That is, if more colors are defined than the number of colors measured when the color conversion table is created, the ink amount data corresponding to the corrected color is calculated more accurately with reference to the color conversion table. be able to. Compared with the color conversion in the color conversion table, the ink amount data corresponding to the corrected color can be calculated more accurately, so that the color conversion accuracy in the color conversion table can be reliably improved by the correction. Become.

  Further, the present invention improves the color conversion accuracy for a specific color by correcting the specific ink amount data. Therefore, when there is a specific color that requires color conversion with higher accuracy than the color conversion accuracy in the color conversion table created in advance, ink amount data corresponding to the specific color or a color close to the specific color is obtained. The specific ink amount data described above is used. Various colors can be adopted as the specific color. For example, a plurality of achromatic colors having different brightness values can be adopted as the specific color. If these colors are specified colors, it is possible to improve the gray color conversion accuracy, and it is possible to print a gray gradation with a very high image quality using the corrected color conversion table.

  In order to improve the color conversion accuracy, the target value may be an ideal value of ink amount data for printing a plurality of colors that are achromatic and have different brightness. Of course, the specific color is not limited to an achromatic color. For example, it may be a color conversion table that performs printing with a skin color and a color close to the skin color as a specific color, and prints with a clean skin color, or a clear sky or sea can be printed with a blue and a color close to blue as a specific color A color conversion table may be used. It is also possible to use a color conversion table that can print beautiful sunsets, etc., with red and colors close to red as specific colors, and with green and colors close to green as specific colors, landscapes including trees can be printed neatly. Various configurations can be adopted, such as a simple color conversion table.

  By the way, the above-described color conversion table correction apparatus may be implemented independently, or may be implemented together with other methods in a state of being incorporated in a certain device. It is included and can be changed as appropriate. In addition, the above-described method for performing correction based on the difference between the ideal value and the colorimetric value, when proceeding according to a predetermined procedure, naturally has an invention in that procedure. is there. Therefore, the present invention can also be applied as a method, and the invention according to claim 8 has basically the same function. When trying to implement the present invention, a predetermined program may be executed by the color conversion table correction apparatus. The present invention can also be applied as a program thereof, and the invention according to claim 9 has basically the same operation.

  Of course, it is also possible to make the structure described in claims 2 to 7 correspond to the above method and program. Any storage medium can be used to provide the program. For example, a magnetic recording medium or a magneto-optical recording medium may be used, and any recording medium that will be developed in the future can be considered in the same manner. In addition, even when a part is software and a part is realized by hardware, the idea of the present invention is not completely different, and a part is recorded on a recording medium and is appropriately changed as necessary. It includes a reading form. Furthermore, the duplication stages such as the primary replica and the secondary replica are exactly the same.

  Furthermore, it can be said that the present invention is also used in a print control apparatus as in claim 10 which executes printing with reference to the color conversion table corrected by the present invention. Of course, it can be said that the present invention is used even for a method or program for realizing a print control apparatus.

Here, embodiments of the present invention will be described in the following order.
(1) Outline of the Present Invention (2) Configuration of Color Conversion Table Correction Device (3) Color Conversion Table Correction Processing (4) Smoothing Processing (4-1) Evaluation Function _En
(4-2) Evaluation function E _d
(4-3) Evaluation function E _q
(4-4) Evaluation function E _v
(4-5) Evaluation function E _i
(4-6) Flow of smoothing process (5) Other embodiment

(1) Outline of the Present Invention FIG. 1 is an explanatory diagram schematically illustrating a color conversion table correction process according to the present invention. In the present embodiment, correction is performed on color conversion table data (LUT 15a) created in advance. The LUT 15a is sRGB data indicating RGB (red, green, blue) gradation values in the sRGB color space and CMYKlclm data indicating gradation values of ink colors of CMYKlclm (cyan, magenta, yellow, black, light cyan, light magenta). This defines the correspondence with (ink amount data). In other words, the LUT 15a is look-up table data that is referred to when image data specifying the color of an image with sRGB data is input and printing is performed by a printing apparatus equipped with CMYKlclm color inks.

  In the example shown in FIG. 1, each component of the sRGB data is represented by a gradation value of 0 to 255, and the gradation value generated by combining the obtained values by dividing the gradation value range of each component into approximately 16 equal parts. Are registered in the LUT 15a. That is, CMYKlclm data for outputting the same color as sRGB data in which each component of RGB has any value of “0, 16, 32,..., 255” is associated. Yes. In the example shown in FIG. 1, it is possible to adopt a configuration in which each component is expressed by gradation values from 0 to 255 and the output color is expressed by combining the gradation values of each component in CMYKlclm data.

The sRGB data is data conforming to the sRGB standard, and the color corresponding to each combination of RGB values is known. That is, arbitrary sRGB data can be associated with data in a device-independent color space such as an L ^* a ^* b ^* space by a known arithmetic expression or the like. However, the CMYKlclm data is a device-dependent color, and the color corresponding to each combination of CMYKlclm values depends on the model of the printing apparatus, the printing conditions, the printing medium, and the like.

  Therefore, in general, a plurality of patches are printed using CMYKlclm data while specifying a printing device, printing conditions, printing medium, and the like to which the LUT 15a is applied, and an output color in the CMYKlclm data is converted to a device-independent color space by a colorimeter or the like. Specified as data in Then, by associating the color indicated by the sRGB data with the color indicated by the CMYKlclm data in the device-independent color space, the LUT 15a is created by associating the CMYKlclm data for outputting the same color as the specific sRGB data. .

Here, the number of colors defined in LUT15a is not particularly limited, is generally determined as described above, it is often a 17 ^three or more. It is very difficult and practically impossible to perform printing and color measurement on a patch having approximately the same number of colors as such an enormous number of colors. Therefore, it is usual to print CMYKlclm data corresponding to the sRGB data by performing an interpolating operation from the colorimetric results by printing a smaller number than 17 ³ , for example, 10 ³ patches.

  However, the color conversion accuracy in the LUT 15a created in this way includes errors due to the above-described interpolation calculation, errors during colorimetry, and the like. For specific colors that require very high-accuracy color conversion, Such an error becomes a factor of image quality degradation. As described above, the color indicated by the sRGB data is known, but even if the CMYKlclm data is associated with the color indicated by the sRGB data so as to output exactly the same color, it cannot be said that the image quality is the best. In some cases. For example, since the output color in the printing apparatus is always influenced by the background color of the print medium, in a specific case such as a bluish background color, printing is performed with a color different from the color indicated by the sRGB data. In some cases, printing results that are preferable are obtained. In this case as well, if the accuracy of color conversion is low, the desired output color that has been changed as described above cannot be obtained, and this causes a sense of discomfort when visually recognizing the image, leading to deterioration in image quality.

The present invention corrects the LUT 15a so that the specific color as described above can be converted to a desired color with high accuracy. In the example shown in FIG. 1, gray is used as the specific color. Yes. That is, the CMYKlclm data associated with the gray defined in the LUT 15a (R = G = B data in the sRGB data, gray data 15d) is corrected. In the present invention, target data 15b in which the ideal value of the data indicating gray is expressed in the L ^* a ^* b ^* space is prepared in advance, and the output color of the CMYKlclm data is the color defined in the target data 15b. Make corrections so that The target data is a predetermined L ^* a ^* b ^* value. However, when the ideal gray value is completely achromatic, an operation to create an L ^* a ^* b ^* value is performed in advance. Even if not, it can be calculated and obtained from the sRGB data.

In the correction according to the present invention, a patch is printed with the CMYKlclm data of the gray data 15d described above, and the print result is measured to correct the CMYKlclm data so that the color difference ΔE between the two is smaller than a predetermined threshold value α. That is, in FIG. 1, the colorimetric value is shown as a vector M (L ^* a ^* b ^* space vector), the L ^* a ^* b ^* value of the target data is shown as a vector T ₀ , and the magnitude of the difference vector between them is shown. If is not less than α, CMYKlclm data is corrected.

In this correction, the L ^* a ^* b ^* value of the patch printed by the CMYKlclm data is moved in a direction to cancel the difference between the L ^* a ^* b ^* value of the target data 15b and the colorimetric value before and after the correction. Is implemented as follows. That is, when a difference is generated between the vector M and the vector T ₀ , it is predicted how the L ^* a ^* b ^* value corresponding to the CMYKlclm data will be changed, and the difference is eliminated. Then, the predicted L ^* a ^* b ^* is calculated.

After predicting the L ^* a ^* b ^* values, with reference to the correspondence data 15c, it calculates the CMYKlclm data corresponding to the L ^* a ^* b ^* value after prediction by interpolation. The correspondence relationship data 15c is data in which CMYKlclm data and the colorimetric values thereof are associated with each other for gray and surrounding colors, and both are associated with a larger number of colors than the gray data 15d.

Of course, the interpolation calculation referring to the correspondence data 15c also includes an error, but in the correspondence data 15c, more colors are defined around the gray than the LUT 15a, so that it corresponds to L ^* a ^* b ^* . CMYKlclm data to be calculated can be calculated with higher accuracy. In the present invention, the correction is repeated until the color difference ΔE becomes smaller than the threshold value α as described above. Therefore, in the end, the difference between the print color by the CMYKlclm data and the target becomes very small. As a result, it is possible to obtain an LUT 15a that can perform color conversion with high accuracy for gray and surrounding colors.

(2) Configuration of Color Conversion Table Correction Device A configuration for carrying out the present invention will be described in detail below. FIG. 2 is a block diagram showing a configuration of the computer 10 functioning as a color conversion table correction apparatus according to the present invention. In FIG. 2, the computer 10 can be realized by a general-purpose personal computer, and is connected to a display 11, a printing device 12, a keyboard 13a, and a mouse 13b. In the present embodiment, the color of the image displayed on the display 11 is specified by the sRGB data, and the color of the image printed by the printing device 12 is specified by the CMYKlclm data.

  The computer 10 includes a program execution environment such as a CPU, RAM, and ROM (not shown), and can execute a PRTDRV 21, the LUT correction program 30, an application program (not shown), and the like under the control of an OS (not shown). A user of the computer 10 can perform an operation input on the LUT correction program 30 or an application program by operating the keyboard 13a, the mouse 13b, or the like. In the PRTDRV 21, an image is printed by controlling the printing device 12 in accordance with a print instruction for an image or the like created by an application program (not shown). In the present invention, it is also possible to receive patch data in which the color of a pixel is specified by CMYKlclm data from the patch data generation unit 31 of the LUT correction program 30 and cause the printing apparatus 12 to execute printing.

  That is, the PRTDRV 21 can implement functions provided by a normal driver for controlling the printing apparatus 12, such as a color conversion function, a halftone processing function, and a function for generating print data to be transferred to the printing apparatus 12. It is. The print data generated by the PRTDRV 21 is transferred to the printing apparatus 12 via the USB I / O 14. That is, the printing apparatus 12 is connected to the computer 10 via USB, and when receiving the print data, records the ink on the print medium according to the print data and forms an image.

  The LUT correction program 30 includes a patch data generation unit 31, a colorimetric value acquisition unit 32, an ink amount data correction unit 33, and a smoothing processing unit 34. Refer to the LUT 15a, target data 15b, and correspondence data 15c. Thus, the correction of the LUT 15a can be performed. In the example shown in FIG. 2, the LUT 15a, the target data 15b, and the correspondence data 15c are recorded in advance in the HDD 15 provided in the computer 10.

  The patch data generation unit 31 is a module that acquires ink amount data, generates the patch data based on the ink amount data, and outputs the patch data to the PRTDRV 21. That is, the patch data generation unit 31 acquires specific ink amount data, and generates patch data indicating a patch having a predetermined area. When this patch data is output to the PRTDRV 21, a patch is printed, and this patch becomes a color measurement target. Here, the ink amount data acquired by the patch data generation unit 31 is recorded in the ink amount data defined in the LUT 15a, the ink amount data generated by the correction by the ink amount data correction unit 33, and the correspondence data 15c. This is ink amount data for specifying a color.

  Of the ink amount data acquired by the patch data generation unit 31, the ink amount data defined in the LUT 15a is the ink amount data defined in the LUT 15a and the correction target, and is shown in FIG. In the example, CMYKlclm data is associated with gray sRGB data. The ink amount data generated by the correction by the ink amount data correcting unit 33 is ink amount data generated at each repetition stage when the correction by the ink amount data correcting unit 33 is repeated as will be described later. The ink amount data for specifying the color recorded in the correspondence relationship data 15c is the ink amount for specifying the color of the patch to be measured when the correspondence relationship data 15c is created as will be described later.

The colorimetric value acquisition unit 32 is a module that acquires the colorimetric values output from the colorimeter 40 via the USB I / O 14. That is, the color meter 40 measures the color of the patch printed by the printing device 12 using the patch data generated by the patch data generation unit 31. In the present embodiment, the colorimeter 40 outputs a color measurement result as an L ^* a ^* b ^* value, and data indicating the L ^* a ^* b ^* value is input to the computer 10 via the USB I / O 14.

  The colorimetric value of the patch printed by the ink amount data specifying the color recorded in the correspondence relationship data 15c is associated with the ink amount data and becomes correspondence relationship data 15c. The color measurement value of the patch printed by the ink amount data defined in the LUT 15a and the ink amount data generated by the correction by the ink amount data correction unit 33 is transferred to the difference vector calculation unit 33a of the ink amount data correction unit 33. It is. This colorimetric value corresponds to the vector M shown in FIG.

The ink amount data correction unit 33 includes a difference vector calculation unit 33a, a correction vector calculation unit 33b, and an interpolation calculation unit 33c. The difference vector calculation unit 33a acquires a colorimetric value from the colorimetric value acquisition unit 32 as described above, and indicates a difference vector (vector M-vector) indicating a difference from the predicted L ^* a ^* b ^* value (vector T). T) is calculated. The ink amount data correction unit 33 repeats the correction as described above. In the first correction for a certain correction target, the ideal value of the target data 15b recorded in the HDD 15 (the vector T shown in FIG. (Corresponding to ₀ ) is the vector T, and the L ^* a ^* b ^* value after prediction is the vector T in the second and subsequent corrections. Further, the difference vector calculated at each correction stage is recorded in a memory (not shown).

  Of course, since the colorimetric value is obtained for a patch printed with the ink amount data defined in the LUT 15a or data obtained by correcting the ink amount data, a target corresponding to the original ink amount data to be corrected is extracted. Thus, a difference vector is calculated for each color to be corrected. That is, the ideal value corresponding to the ink amount data defined in the original LUT 15a is extracted from the target data 15b indicating the ideal values of a plurality of colors, and the difference vector is calculated.

The correction vector calculation unit 33b is a module that extracts the difference vector and target data (vector T ₀ ) to be corrected and calculates a correction vector C. In the present embodiment, the vector indicating the L ^* a ^* b ^* value after the prediction is expressed as vector T ₀ + vector C. The L ^* a ^* b ^* value after the prediction is input to the interpolation calculation unit 33c.

The interpolation calculation unit 33c is a module that performs interpolation calculation with reference to the correspondence data 15c, and calculates ink amount data corresponding to the input L ^* a ^* b ^* value after prediction. The calculated ink amount data is ink amount data generated by the correction by the ink amount data correcting unit 33 described above. Therefore, one correction is performed by the above processing. The ink amount data correction unit 33 can repeatedly perform such correction. In this repetition process, when the magnitude of the difference vector becomes equal to or less than the threshold value α, the correction is finished, The LUT 15a is overwritten with ink amount data corresponding to the predicted L ^* a ^* b ^* value.

  In the embodiment shown in FIG. 2, further correction can be performed on the LUT 15a overwritten as described above. As a further correction, the smoothing processing unit 34 is a module that performs correction so that a change in output color due to a plurality of correction target ink amount data becomes smooth. The smoothing processing unit 34 obtains ink amount data to be corrected and ink amount data that outputs a color close to the output color at the ink amount from the LUT 15a, and evaluates whether the change in the output color is smooth. An evaluation function is calculated.

  Then, the process of calculating the value of the evaluation function by changing the ink amount data to be corrected is repeated to converge the evaluation function to a small value. This evaluation function is calculated by obtaining a predetermined smoothing parameter 15e. The smoothing processing unit 34 only needs to acquire ink amount data having a small evaluation function value by varying the ink amount data, and various algorithms can be employed. Details of this processing will be described later.

(3) Color Conversion Table Correction Processing Next, the correction processing executed in the above color conversion table correction device will be specifically described along the flowchart shown in FIG. The flow shown in the figure is a processing flow of the LUT correction program 30, and the processing shown in FIG. 3 is executed in a state where the LUT 15a and the target data 15b are created in advance and recorded in the HDD 15. In this process, correspondence data 15c is created before correction (step S100).

  In step S100, the patch data generation unit 31 generates patch data. This patch data is a specific color such as gray and has more colors than the gray color defined in the LUT 15a. FIG. 4 shows a processing example when specifying the patch data. FIG. 4A is a diagram showing the sRGB data defined in the LUT 15a in an orthogonal RGB space. That is, since the value range of the sRGB data is the same for each color component value, the values that the sRGB data can take form a cube indicated by a broken line in the orthogonal RGB space.

  Arbitrary sRGB data is represented by lattice points that exist in the cube and each color component value is an integer. In this cube, gray data corresponds to grid points on a straight line KW connecting the origin K and the vertex W of the cube. In the LUT 15a, sRGB data and corresponding CMYKlclm data obtained by combining values obtained by dividing each color component value range into approximately 16 parts and corresponding CMYKlclm data are registered, so there are 17 grid points including both ends on the straight line KW.

  FIG. 4 shows a state in which a cube in the RGB space is cut to form a plane P including a straight line KW in order to explain how the number of grid points in the RGB space increases around the 17 lattice points. . FIG. 4B shows a state in which the plane P is projected onto the RG plane and the lattice points on the plane P are connected by straight lines parallel to the R axis and the G axis. In the operation in step S100, lattice points are extracted more densely than the lattice points shown in FIG. 4B on and around the straight line KW.

  FIG. 4C is an enlarged view showing the vicinity of both ends of the straight line KW in FIG. 4B. In FIG. 4C, grid points corresponding to the sRGB data registered in the LUT 15a are indicated by black circles, and straight lines connecting the grid points when each color component value range is divided into approximately 52 parts are indicated by solid lines. That is, FIG. 4C shows an example in which sRGB data corresponding to lattice points when each color component value range is divided into approximately 52 equal parts in step S100 is extracted on and around the straight line KW.

  If the grid points are extracted from the periphery of the straight line KW in this way, the grid points on the straight line KW and within a certain distance from the straight line KW are extracted and used as sRGB data for creating the correspondence data 15c. Good. FIG. 4C shows an example in which grid points within a broken line at a certain distance from the straight line KW are extracted. For example, white circles are extracted on the R axis.

  When a grid point finer than the grid point corresponding to the sRGB data registered in the LUT 15a is extracted, the ink amount data corresponding to the grid point is calculated with reference to the LUT 15a. The patch data generation unit 31 creates patch data specifying a color based on the calculated ink amount data, and outputs the patch data to the PRTDRV 21. As a result, the printing apparatus 12 prints a plurality of patches. When the patch is printed, colorimetry is performed by the colorimeter 40, and the colorimetric value is acquired by the colorimetric value acquisition unit 32. Since the ink amount data used for printing the patch is known in advance, the ink amount data is associated with the colorimetric value. The data created as a result is the above-described correspondence data 15c.

  In the printing and colorimetry operations described above, the same operation is performed for the black circle lattice points on the straight line KW, that is, the lattice points corresponding to the sRGB data registered in the LUT 15a. The colorimetric value obtained by this work is used as the initial value of the vector M described above. In creating the correspondence data 15c, sRGB data corresponding to lattice points when each color component value range is divided into approximately 52 parts is considered. However, when printing and colorimetry are actually performed, straight lines are considered. Since only KW and surrounding grid points are extracted, the number of patches does not increase so that printing and colorimetry operations cannot be performed.

As described above, when the creation of the correspondence data 15c and the initial value acquisition process of the vector M are performed in the process of step S100, the correction process of the LUT 15a is performed in and after step S105. For this purpose, first, a number N indicating the number of corrections is initialized (step S105). Then, an L ^* a ^* b ^* value that will eliminate the difference between the ideal value and the colorimetric value is predicted according to the following equation (1) (step S110).

Here, the vector T _{N + 1} represents the number of corrections N + 1, and the vector T _i represents the L ^* a ^* b ^* value after prediction at the correction number i. The vector T ₀ is the target data 15b as described above. L ^* a ^* b ^* values registered in The vector M _i is the colorimetric value of the patch at the correction number i, while the vector M ₀ is the colorimetric value of the patch printed with the ink amount data registered in the LUT 15a.

The expression (1) shown in FIG. 3 is an expression for performing the (N + 1) th prediction in consideration of the difference before the Nth correction, and the correction of the second term with respect to the target vector T ₀ shown in the first term. This is a formula for performing prediction by adding the vector C. Here, the correction vector C is calculated from the difference vector (vector M-vector T) between the colorimetric value of the patch and the ideal value or the corrected value (predicted value) before the (N-1) th time. That is, the difference vector calculation unit 33a acquires the colorimetric value and the target data 15b or the corrected L ^* a ^* b ^* value before the N-1th time to create a difference vector, and the correction vector calculation unit 33b C is calculated, thereby calculating the value of the above equation (1).

  For example, in the first prediction, the inverse vector of the difference vector (vector M-vector T) between the colorimetric value of the patch and the target data 15b becomes the correction vector C. In the second and subsequent times, the inverse of the difference vector in the second term. A correction vector C is obtained by multiplying the vector by a weight corresponding to the reciprocal of the magnitude of the previous difference vector and adding it. Such a correction vector is a difference vector indicating a difference between the ideal value and the colorimetric value, or a difference between the colorimetric value and a value in a device-independent color space indicating a color corrected by the ink amount data correcting unit. It can be said that the difference vector is calculated from a vector having a reverse component of the difference vector.

After performing the prediction as described above, the interpolation calculation unit 33c refers to the correspondence data and calculates ink amount data corresponding to the obtained predicted vector T _{N + 1} (step S115). The ink amount data calculated here is the corrected ink amount data. When the ink amount data is calculated, the patch data generation unit 31 acquires the calculated ink amount data, generates patch data, and outputs the patch data to the PRTDRV 21 (step S120). As a result, the printing apparatus 12 prints the patch whose color is specified by the corrected ink amount data.

When the patch is printed, the colorimetric value acquisition unit 32 acquires a colorimetric value that is a colorimetry result obtained by the colorimeter 40 (step S125). Here, the acquired colorimetric value is an accurate value of the color output based on the corrected ink amount data, and is expressed as a vector M _{N + 1} . Therefore, in order to compare the colorimetric value with the ideal value of the color to be output by the above-described target data 15b, that is, the converted ink amount data referring to the LUT 15a, a difference vector (vector T ₀ -vector M _{N + 1).} ) Is calculated. Then, it is calculated whether or not the magnitude of the difference vector is equal to or smaller than a predetermined threshold value α (for example, color difference 1) (step S130).

  If it is not determined in step S130 that the magnitude of the difference vector is equal to or smaller than the predetermined threshold value α, the number N indicating the number of corrections is incremented (step S135), and steps S110 to S130 are repeated. If it is determined in step S130 that the magnitude of the difference vector is equal to or smaller than the predetermined threshold value α, the color printed with the corrected ink amount data and the color registered in the target data 15b are very close. The LUT 15a is updated with the corrected ink amount data (step S140). Then, it is determined whether or not the update has been completed for all the ink amount data to be corrected (step S145), and the processes in and after step S105 are repeated until it is determined that the update has been completed for all.

FIG. 5 is an explanatory diagram for explaining how correction is performed in steps S110 to S130 described above. This figure shows a state in which coordinates in the L ^* a ^* b ^* space are projected onto the a ^* b ^* plane. Also, the coordinates specified by the vector T are shown as black circles, and the coordinates specified by the vector M are shown as white circles. A vector T ₀ indicating the color value registered in the target data 15b is an ideal value, and is a fixed value for each ink amount data to be corrected.

If the colorimetric value of the patch printed with the ink amount data (pre-correction data) registered in the LUT 15a is the vector M ₀ , the predicted value in the first correction is the target vector T ₀ and the difference vector. The inverse vector (vector T ₀ -vector M ₀ ) is added. That is, there is a difference corresponding to the difference vector between the target and the colorimetric value. Therefore, if correction corresponding to the inverse vector of the difference vector is performed on the ink amount data, the corrected ink amount data This is based on the idea that the print result at the target approaches the target.

Actually, there may be a case where the target and the colorimetric value are not very close by one correction due to an interpolation error, a colorimetric error, or the like. In FIG. 5, the vector M ₁ indicating the colorimetric value of the patch printed with the ink amount data corresponding to the color indicated by the predicted vector T _{1 in the first} correction matches the target vector T _0. Instead, an example in which a difference occurs is shown. In this embodiment, the correction is repeated a plurality of times in order to eliminate such a difference, but at this time, the prediction is performed in consideration of the difference vector in the past correction, and this correction is formulated. Is the above formula (1).

That is, the second term of Equation (1) is a linear combination of the inverse vector of the difference vector at each correction stage multiplied by a weighting coefficient that increases as the distance from the target to the colorimetric value decreases. . More specifically, in order to calculate the vector T ₂ in FIG. 5, the difference vector (vector M ₀ - vector T ₀₎ - a distance d ₁ (vector T ₀ vector M ₁₎ the distance d ₀ and difference vector Calculate and express a larger value as the distance is shorter by the reciprocal of both. Each reciprocal is normalized by a value obtained by adding the reciprocal of each distance ((1 / d ₀ + 1 / d ₁ ) in FIG. 5), so that each coefficient does not exceed 1.

According to this equation, the closer the point indicated by the colorimetric value vector _MN and the point indicated by the target vector T ₀ are, the greater the inverse vector of the difference between them will contribute to the calculation of the correction vector C. it can. In other words, the closer the colorimetric value is to the target, the more likely it is to be able to accurately predict the output color based on the corrected ink amount data, so the closer the colorimetric value is to the target, the greater the contribution of the difference vector. Thus, the output color based on the ink amount data can be brought close to the target quickly.

On the other hand, the reason for considering other difference vectors without considering only the difference vector with the smallest difference is to prevent vibration and divergence during the repetition of correction. That is, when only the difference between the color after the N-th correction and the colorimetric value for the ink amount data is taken into consideration in the N-th correction, vibration and divergence can occur as described above. For example, when the colorimetric value corresponding to the ink amount data corresponding to the vector T _N in FIG. 5 is substantially the same as the vector M _M (M is a natural number smaller than N), the same correction is performed without considering the past history. Will be repeated.

  In addition, when the difference is increased by the N-th correction due to an interpolation error, a colorimetric error, or the like, the influence of the N-th correction is large unless the past history is taken into consideration. However, this causes a situation where the colorimetric value does not approach the target. However, in the present embodiment, since the above linear combination is performed including not only the difference vector in the Nth correction but also the difference vector in the previous correction, the above vibration and divergence can be effectively prevented. The output color based on the ink amount data can be brought close to the target quickly.

  After updating the ink amount data to be corrected as described above, the LUT 15a that can perform color conversion with high accuracy on the ink amount data to be corrected and the surrounding colors is obtained. In terms of color conversion accuracy, it is possible to perform very high-accuracy color conversion with the LUT 15a at this stage, but further, the relationship between the ink amount data to be corrected is adjusted to perform higher-quality printing. It can also be possible.

  That is, although the above-described correction enables highly accurate color conversion, the relationship between the correction target ink amount data is not taken into consideration. Therefore, further correction may be performed so that the gradation of the output color to be corrected changes smoothly. In the flowchart shown in FIG. 3, the above-described smoothing processing unit 34 performs such correction (smoothing process: step S150), and the correction target ink amount data in the LUT 15a is updated with the data after the smoothing process, and the HDD 15 is updated. (Step S155) to complete the color conversion table correction process.

(4) Smoothing Processing FIG. 6 is an explanatory diagram for explaining an overview of processing in the smoothing processing unit 34. The left side of the figure shows the gray data 15d after the correction in step S140, and the smoothing processing unit 34 performs a smoothing process on the corrected data. In this smoothing process, defining an evaluation function E _s for the ink amount data after correction (CMYKlclm data), the ink amount data by changing the ink amount data so as to minimize the evaluation function E _s Smoothes the output gradation by.

That is, to optimize the amount of ink in the ink in the ink amount data so as to minimize the evaluation function E _s. Various algorithms can be adopted for this optimization processing, but since the ink amount is usually an integer value, an algorithm applied to the integer value is preferable.

The evaluation function E _s is a linear combination of a plurality of evaluation functions (E _q , E _v , E _i ) for evaluating the change in gradation indicated by the ink amount data. In this embodiment, the evaluation functions E _n and E _d are defined, and when the evaluation function E _s is optimized to be as small as possible, the constraint conditions are set so that these evaluation functions E _n and E _d do not become larger than zero. Imposing. The function forms of these evaluation functions are determined in advance, and are calculated by referring to the parameters described in the smoothing parameter 15e. Hereinafter, each evaluation function will be described in detail.

ここで、Ｊはインク数を示す自然数であり、ｎは色を識別するための識別番号を示す自然数である。また、ｍはスムージング対象のインク量データを示す自然数であり、上記グレーデータ１５ｄに規定されたインク量データを識別する番号である。従って、ｉ_mnは、ｍ番目のインク量データにおけるインク色ｎのインク量を表す。ｔｍｐ_nはインク色番号ｎの場合の一時変数である。 (4-1) Evaluation function _En
Evaluation function E _n is the evaluation function for evaluating whether the ink amount is a negative value, defined as the following equation (2).

Here, J is a natural number indicating the number of inks, and n is a natural number indicating an identification number for identifying a color. Further, m is a natural number indicating the ink amount data to be smoothed, and is a number for identifying the ink amount data defined in the gray data 15d. Therefore, i _mn represents the ink amount of ink color n in the m-th ink amount data. tmp _n is a temporary variable for ink color number n.

_Wn is a weighting factor, and is specified in advance as the smoothing parameter 15e. Equation (2) is multiplied by a weighting coefficient w _n with respect to the square of its magnitude when the ink amount is negative is obtained by linear combination from this evaluation function E _n "0" in this embodiment Apply restraint conditions so as not to grow. Therefore, it is possible to vary upon optimization of the amount of ink the evaluation function E _s in the smallest possible value, the amount of ink so that the ink amount does not become a negative value.

ここで、Ｋはインク量制限の条件数を示す自然数であり、ｌはその識別番号を示す自然数であり、Ｊ，ｎ，ｍ，ｉ_mnは上記式（２）と同様である。Ｄｌは条件番号ｌのインク量を表し、ｕ_lnは条件番号ｌの場合にインク色番号ｎのインク量を加算するか否かを決定するための係数で、”０”あるいは”１”を取る。 (4-2) Evaluation function E _d
The evaluation function E _d is an evaluation function for evaluating whether or not the ink amount recorded on the print medium is equal to or less than the ink amount limit, and is defined as the following equation (3).

Here, K is a natural number indicating the condition number for limiting the ink amount, l is a natural number indicating the identification number, and J, n, m, and i _mn are the same as in the above equation (2). Dl represents the ink amount of the condition number l, and u _ln is a coefficient for determining whether or not to add the ink amount of the ink color number n in the case of the condition number l, and takes “0” or “1”. .

For example, if the condition number l is a condition relating to the ink amount restriction of the primary color (single color) of ink color number n = n1, u _ln = 1, and u _ln = 0. If condition number 1 is a condition related to the total ink amount, u _ln = 1 for all ink color numbers. D _Ll represents the ink amount limit value for the condition number l. That is, since there may be a plurality of conditions for limiting the amount of ink depending on the number of ink colors used at the same time and the combination of inks, each condition can be identified by being numbered.

Note that tmp _l is a temporary variable for the condition number l. W _d is a coefficient representing a weight. These w _d , u _ln , and D _Ll are specified in advance as the smoothing parameter 15e. In the above equation (3), when the ink amount (Dl) exceeds the ink amount limit value (D _Ll ), the temporary variable tmp _l becomes a value larger than “0”, and the linear combination is the evaluation function E. _d . In this embodiment, since applying the evaluation function E _d is "0" increases become not so constrained conditions than, upon optimization of the amount of ink the evaluation function E _s in the smallest possible value, so as not to exceed the ink amount limiting value The amount of ink can be varied.

ここで、Ｊ，ｎ，ｍは上記式（２）と同様であり、ｉ_qはインク量の２次微分ベクトルの各要素を示している。ｗ_qは重み係数であり、各インク量の２次微分値が評価関数Ｅ_qに寄与する度合いを決定する値である。 (4-3) Evaluation function E _q
The evaluation function E _q is an evaluation for evaluating the change rate of the ink amount in the adjacent ink amount data (data adjacent when the sRGB data is arranged in the order of size like the gray data 15d shown in FIG. 6). It is a function and is defined as the following formula (4).

Here, J, n, and m are the same as in the above equation (2), and i _q represents each element of the secondary differential vector of the ink amount. w _q is a weighting factor, and is a value that determines the degree to which the secondary differential value of each ink amount contributes to the evaluation function E _q .

The second-order differential vector of the ink amount can be calculated by various methods. As an example applied to the present embodiment, the first-order differential as shown in the formula (5) is used to express 2 as shown in the formula (6). A configuration for calculating the second derivative can be employed.

Here, m is a natural number indicating the ink amount data to be smoothed similarly to the above formula (2), and here, the ink amount data to be smoothed is T + 1 (numbers 0 to T). R _m represents the tone value of sRGB data, vector I _m represents a vector whose elements a gradation value of each color of ink amount data.

In the present embodiment, since sRGB data to be smoothed is R = G = B, the sRGB data is not a vector quantity but a scalar quantity. Here, the number m is assigned in ascending order of the sRGB data values, and R ₀ , R ₁ ,..., R _T are assigned in ascending order of gradation values, and the ink amount vector corresponding to each gradation value is the vector I _0. , I ₁ ,..., I _T. For example, in FIG. 6, the ink amount data with the value “0” of the sRGB data corresponds to the ink amount data with the number 0, and the ink amount data with the value “255” of the sRGB data corresponds to the ink amount data with the number T. .

In the first equation of equation (5), the vector I _pm, since it shows a change in ink amount vector I _m with respect to the change of the gradation value R _m, which is first-order differential of the ink amount data, i.e., the ink amount The rate of change of the vector is shown. The second equation of the formula (5) defines a tone value R _pm of sRGB data comprising such a first-order differential value, as shown in the second equation, the tone value R _m in the present embodiment And the midpoint of the gradation value R _{m + 1} .

When the primary differential vector I _pm and its gradation value R _pm are defined as described above, it is possible to calculate the secondary differential vector by the following equation (6).

Here, m is the same as the above formula (5), but its value range is 1 to T-1. In the above equation, w _q is specified in advance as the smoothing parameter 15e. In the equation (6), the vector I _qm indicates the change of the primary differential vector I _pm with respect to the change of the gradation value R _pm , so this is the second derivative of the ink amount data, that is, the change of the ink amount vector. It shows the rate of change of rate. If the change rate of the change rate of the ink amount vector can be evaluated, the smoothness of the gradation change based on the ink amount data can be evaluated.

  That is, when the change rate of the ink amount changes abruptly, a tone jump in which the gradation of the output color based on the ink amount data changes rapidly is likely to occur, but the change rate of the ink amount changes smoothly. If so, such a tone jump is unlikely to occur. FIG. 7 is an explanatory diagram for explaining the basic concept of evaluation based on such secondary differentiation based on an example. In the figure, the relationship between the brightness of the output color expressed by dark ink (C or M ink) and light ink (lc or lm ink) and the ink recording rate is shown.

The solid line in the figure shows an example in the ink amount data before the smoothing process. When the lightness value range that can be expressed is 0 to 100, the dark ink recording rate decreases as the lightness becomes darker from 0. However, light ink starts to be used at a certain lightness, dark ink is not recorded at a certain lightness, and only light ink is recorded at lighter brightness than that, and the lightness is gradually increased. In the example shown in the figure, a tone jump is likely to occur between the lightness L ₀ for switching the use / non-use of light ink and the lightness L ₁ for switching the use / non-use of dark ink.

As in the example shown by the solid line, when the light ink is not used at lightness L ₀ or less and the amount of light ink used increases linearly at lightness L ₀ or more, the second derivative of the light ink amount at lightness L ₀ The absolute value increases. Further, without using the dark ink in the lightness L ₁ or more, when the amount of dark ink in lightness L ₁ or less linearly increases, the absolute value of the secondary differential value of the dark ink in the lightness L ₁ increases. Of course, not only in these examples, but also if the rate of change in the ink amount changes abruptly, the absolute value of the secondary differential value increases. Also, not only when using dark and light inks, but also when inks with different densities are not used, even when three types of ink density are used, the rate of change of the ink amount changes abruptly depending on the secondary differential value. Then, the absolute value of the secondary differential value becomes large.

In the evaluation function E _q , as described above, the square of each element in the secondary differential vector is multiplied by the weighting coefficient w _q and linearly combined. Therefore, the value of the evaluation function E _q increases as the secondary differential value increases. In this embodiment, in the definition of such an evaluation function, varying the ink amount data so as to minimize the evaluation function E _s. Therefore, there is a high tendency to obtain ink amount data with a small second order differential value, and ink recording as indicated by a one-dot chain line in FIG. Ink amount data is obtained such that the rate changes smoothly. Accordingly, it is possible to obtain ink amount data in which tone jump is unlikely to occur.

(4-4) Evaluation function E _v
Evaluation function E _v is whether the difference between the output color by the ink amount data after alternative output color by the original ink amount data when replacing ink amount of a certain color in alternative color inks is large An evaluation function for evaluation. As this evaluation function, for example, a process for replacing a minute amount of a certain ink with a replaceable ink can be defined by a function formulated. In other words, when the ink is replaced with a replaceable ink, the smoothness of the gradation change in the ink amount data is improved. For example, when the dark ink amount 1 and the light ink amount 4 are output colors having substantially the same lightness, if the ink amount data is changed so that the dark ink is decreased by a small amount by 1 and the light ink is increased by a small amount by 4, The gradation change can be smoothed with almost no change in the output color based on the ink amount data.

ここで、ｍは上記式（２）と同様であり、ｒはスムージング前のインク量データであることを示す識別符号であり、ｗ_vは補正前後のインク量をマトリクス変換した場合の差分が評価関数Ｅ_vに寄与する度合いを決定する値である。 More specifically, this evaluation function can be defined as in the following formula (7).

Here, m is the same as the above equation (2), r is an identification code indicating that the ink amount data is before smoothing, and w _v is an evaluation of the difference when the ink amount before and after correction is subjected to matrix conversion. It is a value that determines the degree of contribution to the function E _v .

s is an identification code indicating an element after matrix conversion, and v _ms indicates each element of the result of converting the changed ink amount data by the matrix A, and is each element in the following equation (8).

Further, v _rms in Expression (7) indicates each element of the result of converting the ink amount data before fluctuation (before smoothing processing) by the matrix A, and is each element in Expression (9) below.

当該（１０）式に示す例は、ＣＭＹＫｌｃｌｍデータを仮想的にＣＭＹデータに変換するマトリクスであり、マトリクスＡの１列目〜６列目までがＣ〜ｌｍに対応している。 In the above formula, w _v is specified in advance as the smoothing parameter 15e. In the equations (8) and (9), the matrix A indicates the substitution relationship between the inks that can be substituted by using the substitution ratio between the inks that can be substituted as elements. For example, the following equation (10) The value of can be adopted.

The example shown in the equation (10) is a matrix for virtually converting CMYKlclm data into CMY data, and the first to sixth columns of the matrix A correspond to C to lm.

  That is, when converting each color of CMYKlclm to the CMY color space, each color of C, M, and Y ink is multiplied by a coefficient 1 and contributes only to each CMY color, whereas K ink is multiplied by a coefficient 1 Contributes to all CMY colors. This is based on the fact that when an equal amount of each CMY ink is recorded, the color becomes almost achromatic. On the other hand, each color of C and M inks, which are dark inks, is multiplied by a coefficient of 1 to contribute only to each color of CM, while light inks lc and lm are multiplied by a coefficient of 0.25 and only each color of CM is multiplied. Contribute to. This is based on the fact that the output color based on the dark ink amount 1 and the output color based on the light ink amount 4 have almost the same lightness as in the above example.

In the evaluation function E _v shown in the equation (7), the square of the difference between each element in the virtual CMY space obtained by converting the ink amount data before and after correction is multiplied by the weighting factor w _v and linearly combined as described above. ing. Thus, larger difference in each element of the virtual CMY space, evaluation value of the function E _v is increased, the value of the more substantial change color before and after the correction becomes large evaluation function E _v is increased.

In the present embodiment, in the definition of such an evaluation function, varying the ink amount data so as to minimize the evaluation function E _s. As a result, the amount of ink can be varied so that the substantial change in color is small, and it is possible to perform a replacement process for ink that can be replaced without changing the color. FIG. 8 is an explanatory diagram for explaining such an alternative to ink. Also in this figure, the relationship between the brightness of the output color expressed by dark ink (C or M ink) and light ink (lc or lm ink) and the ink recording rate is shown. In addition, the solid line shown in the figure also shows an example of the ink amount data before the smoothing process, as in FIG.

According to the above equation (7), the variation in the ink amount data that allows the ink to be replaced within a small range of the variation in color is allowed. For example, as indicated by the alternate long and short dash line near the lightness L ₀ described above. A state in which the dark ink is reduced and the light ink is increased is included as an ink amount variation option. Further, a state in which the dark ink is increased and the light ink is decreased as shown by a one-dot chain line in the vicinity of the lightness L ₁ described above is included as an ink amount variation option. Of course, the opposite is true, for example, the state in which the dark ink is increased near the lightness L ₀ and the light ink is decreased or the light ink is increased near the lightness L _1. Is also included as an option for ink amount variation.

On the other hand, the evaluation function E _s, so including the evaluation function E _q of the above in conjunction with the evaluation function E _v, in the course of the process to minimize the evaluation function E _s, together reduce the evaluation function E _q an evaluation function E _v . Therefore, in the ink amount data obtained as a result, it is possible to acquire ink amount data that allows the ink recording rate to change smoothly while changing the ink amount by substituting for replaceable ink. Become.

ここで、ｎ，ｍは上記式（２）と同様であり、ｒは上記式（７）と同様である。また、ｗ_inはインク色番号ｎにおいて、補正前後のインク量変化が評価関数Ｅ_vに寄与する度合いを決定する重み係数であり、ｗ_iは全インク色番号についての線形結合が評価関数Ｅ_iに寄与する度合いを決定する重み係数である。以上の式において、ｗ_i、ｗ_inは、上記スムージングパラメータ１５ｃとして予め特定されている。 (4-5) Evaluation function E _i
The evaluation function E _i is an evaluation function for evaluating whether or not the ink amount before correction is largely changed by the correction by the smoothing process, and is defined as the following formula (11).

Here, n and m are the same as the above formula (2), and r is the same as the above formula (7). Also, w _in in ink color number n, is a weighting factor which the ink amount change before and after correction is determined the degree contribute to the evaluation function E _v, w _i evaluation is a linear combination of all the ink colors ID function E _i This is a weighting factor that determines the degree of contribution to. In the above formula, w _i, w _in is specified beforehand as the smoothing parameter 15c.

In equation (11), since the square of the difference between the ink amounts before and after correction is multiplied by the weighting coefficient and linearly coupled, the value of the evaluation function E _i increases as the change in the ink amount before and after correction increases. In the present embodiment, in the definition of such an evaluation function, varying the ink amount data so as to minimize the evaluation function E _s. As a result, the ink amount data can be gradually changed without abruptly changing the ink amount of each color.

In other words, in various optimization algorithms, if the value of a variable is changed suddenly, there are many cases where an appropriate solution cannot be obtained because it falls into a local solution or the solution diverges. If the rapid change in the ink amount data is suppressed by the function E _i , an appropriate solution can be easily obtained. Further, by adjusting the weighting coefficient, it is possible to change the ink amount data while preventing an apparently abnormal value from being generated as the ink amount data.

  For example, the ink amount of the ink color number n is “0” in both the ink amount data numbers m + 1 and m−1, and the ink amount of the ink color number n is also in the smoothing target ink amount data (number m). If the ink amount data of the ink color number n becomes other than “0” by the correction when it is “0” before the correction, there is a high possibility that the smoothness of the gradation is impaired. Therefore, the ink amount data can be considered as an abnormal value in terms of smoothness of gradation.

This example is also shown in FIG. That is, if the ink amount of the dark ink is a finite value at the position indicated as Ab in the figure, it is unlikely that the gradation will be smooth at that point. Therefore, the periphery of the ink amount data is "0" for a is ink color by increasing the value of the weighting coefficient w _in the evaluation function E _i is adjusted to be larger when the ink amount data is changed. As a result, when the ink amount data is changed, it is possible to prevent the ink amount data from changing to an abnormal value.

(4-6) Flow of Smoothing Processing Next, a specific processing example when the ink amount data is optimized using the above evaluation function will be described. This process is performed in step S150 on the data after the process in step S140 shown in FIG. 3 is completed and the correction of the gray data 15d is completed. FIG. 9 is a flowchart showing details of step S150, which is an algorithm called an annealing method. This algorithm simulates the situation when a substance is slowly solidified, and a temperature variable is virtually introduced, and the amount of ink is varied by a random number while gradually decreasing the temperature. And gradually decreasing the evaluation function E _s in the process to optimize the amount of ink by determining the amount of ink to the evaluation function E _s becomes small.

  In this flowchart, the maximum value tmax for selecting the ink amount data to be corrected is determined in advance, the maximum value tmax is selected, and the process is repeated until the smoothing process is completed for each selection. In order to perform this process, first, “0” is substituted into a variable t indicating the number of times of selection of the ink amount data (step S200). The maximum value tmax is previously described in the smoothing parameter 15e.

The ink amount data to be corrected is selected by a uniform random number (step S202). That is, the ink amount data to be corrected is selected at random. Next, an evaluation function E _qb , an evaluation function E _vb , and an evaluation function E _ib are obtained from the ink amount data. Here, the evaluation function E _qb , the evaluation function E _vb , and the evaluation function E _ib are the evaluation function E _q , the evaluation function E _v , and the evaluation function E that are calculated based on the ink amount data at the previous correction stage in the repetitive processing. _i . In step S204, since the first correction is not performed on the ink amount data to be corrected, the evaluation function E _v and the evaluation function E _i taking into account the change before and after the correction are set to “0”, and the evaluation function Only _Eq is calculated.

  Further, a predetermined temperature initial value β is substituted for the above-described temperature variable dt (step S206), and an initial value “0” is substituted for a variable d indicating the number of temperature variable changes (step S208). That is, in the flowchart shown in FIG. 9, the temperature variable dt is updated and gradually reduced in step S242 described later, but the maximum value dmax of the number of updates is determined in advance, and the temperature variable is reduced until this dmax is reached. Repeat the process. The initial temperature value β and the maximum value dmax are described in advance in the smoothing parameter 15e.

  Further, an initial value “0” is substituted for a variable ic indicating the number of times ink amount data has been updated (step S210), and an initial value “0” is substituted for a variable i indicating the number of repetitions of steps S214 to S238 (step S212). Then, the ink amount data update process after step S214 is performed. The maximum value of the variable ic is icmax, and the maximum value of the variable i is imax. In this embodiment, the ink amount data is maximized icmax times while the repetitive process for updating the ink amount data is repeated at most imax times. It is being updated. These maximum values icmax and imax are values previously described in the smoothing parameter 15e.

  In the repetitive processing, first, the ink color to be updated is specified by a uniform random number from the ink amount data selected in step S202 (step S214). When the ink color to be updated is specified, the update amount di is acquired with a normal random number (step S216). That is, the update amount di is determined by a random number such that the distribution of the update amount di follows a normal distribution, thereby mimicking a phenomenon in the natural world. As a result, escape from the local solution is attempted. Since the ink amount data is an integer gradation value, the update amount di is also an integer. The parameters of the normal distribution may be determined from various points of view, for example, a range that does not deviate from the purpose of smoothing the gradation by fine adjustment of the ink amount data, or a range that does not require too many updates to obtain a solution. For example, it is possible to adopt a normal distribution with gradation value dispersion of about “1 to 3”.

Since the update amount di follows a normal distribution, the update amount di may be “0”, but if the update amount is “0”, there is no point in calculating the evaluation function. Therefore, in step S218, it is determined whether or not the update amount is “0”. If the update amount di is “0”, the processing in steps S220 to S236 is skipped. In step S218, when the update amount di is not determined to be "0", the addition of the update amount di in ink amount i _n (step S220). Here, the ink amount i _n is the ink color of the ink amount data selected at step S214. The ink amount i _n0 before correction is also held in a RAM (not shown).

After updating the ink amount, with reference to the weighting factor and the like required parameters recorded on the smoothing parameter 15e, in an ink quantity i _n the updated calculates the evaluation function E _d and the evaluation function E _n (step S222 ). These evaluation functions differ from evaluation functions E _q , evaluation functions E _v , and evaluation functions E _i described later, and other ink amount data, that is, ink amount data other than the ink amount data to be smoothed and the ink amount before correction. Can be calculated without referring to.

The evaluation and the function E _n an evaluation function E _d, instead of minimizing the value of the function, the value of the function as described above is "0" becomes greater than imposing a condition that should not, to act as constraints in optimizing the evaluation function E _s. In the flowchart shown in FIG. 9, one of the evaluation function E _d and these evaluation functions E _n is "0" or determines a larger (step S224). Then, "0" so as not to adopt the ink amount after the update by returning the amount of ink i _n the original value ink amount i _n0 with skips step S226~S236 when it is greater than a determination in step S224 ing.

Step when evaluated either a function E _n an evaluation function E _d is not determined to be greater than "0" in S224, by using the ink amount i _n the updated evaluation function E _qa, the evaluation function E _va, evaluation function _Find E _ia . Here, reference character a was subjected to the evaluation function is the sign for distinguishing the above-described code b, which indicates that a value calculated by the ink amount i _n the updated.

After calculating the evaluation function E _qa , the evaluation function E _va , and the evaluation function E _ia , the difference dE (dE = (E _qa + E _va + E _{ia) between} the evaluation functions before and after the correction is used to determine whether or not the ink amount should be adopted. ) − (E _qb + E _vb + E _ib )) is calculated (step S228). Then, it is determined whether or not the difference dE is smaller than “0” or the uniform random number rnd is smaller than exp (− (dE / dt)) (step S230).

If it is determined in step S230 that the difference dE is smaller than “0” or the uniform random number rnd is smaller than exp (− (dE / dt)), the evaluation function E _qb , the evaluation function E _vb , the evaluation The value of the function E _ib is overwritten with the value of the evaluation function E _qa , the evaluation function E _va , and the evaluation function E _ia , respectively, and the variable ic indicating the number of times the ink amount data has been updated is incremented (step S232). Step S230 in the difference dE is smaller than "0" state or a uniform random number rnd is exp - not when ((dE / dt)) is not determined to be smaller than state, ink amount i _n that as the ink amount after updating Returning to the original ink amount i _n0 as appropriate (step S234).

That is, if less of the more difference dE is "0", indicates that the evaluation function E _s by updating the ink amount becomes smaller, the evaluation function E _s of the gradation by updating the amount of ink that is reduced It means that smoothness has improved. Therefore, in step S232, the updated ink amount i _n is adopted, and the evaluation function E _qb , the evaluation function E _vb , and the evaluation function E _ib are updated.

On the other hand, in certain cases even in the algorithm in this embodiment is simulating a phenomena in nature, temporarily evaluation function E _s is increased, the permissible. As a result, it is possible to escape from the drop in the local solution, and to quickly converge to an appropriate solution. The conditional expression rnd <exp (− (dE / dt)) in step S230 is a formulation of a condition that allows an increase in dE.

  In this equation, the uniform random number rnd is a real number of 0 ≦ rnd ≦ 1, and when the value of the uniform random number is smaller than exp (− (dE / dt)), the ink amount is allowed to be updated. Become. That is, this expression defines the probability that the value exp (− (dE / dt)) is smaller than the uniform random number rnd so as to depend on the difference dE and the temperature dt, and depends on the difference dE and the temperature dt. Allow an increase in dE with some probability.

  More specifically, exp (−dE / dt) is “1” when dE = 0, and is a function that decreases monotonously as dE increases. Therefore, the larger the value of dE, the higher the rnd <exp It can be said that the probability of satisfying (−dE / dt) is reduced. In addition, since the decrease rate of exp (−dE / dt) is larger as the temperature dt is smaller, the probability of satisfying the condition is even lower when the temperature is low even if the temperature is the same dE.

  As described above, after the ink amount is updated in step S232 or the ink amount is not updated in step S234, it is determined whether or not the variable ic indicating the number of times the ink amount data has been updated is equal to or greater than icmax (step). S236). If it is determined in step S236 that the variable ic is greater than or equal to icmax, step S238 is skipped, and if it is not determined in step S236 that the variable ic is greater than or equal to icmax, step S238 is executed.

  That is, when the ink quantity data update count ic exceeds icmax, the process exits the loop of steps S214 to S238. This is a process for preventing the redundant repetition process from being performed when the ink amount data is updated sufficiently many times. For example, when the temperature variable dt described above is large, the ink amount is updated with a high probability, and a sufficient number of times (icmax times) are updated without repeating the loop from step S214 to S238 imax times. There may be. In this case, even if the steps S214 to S238 are repeated further, the possibility of improving the gradation is low and there is no point in performing the repeated processing.

  It is sufficient that icmax can designate a sufficient number of times for updating the ink amount. For example, a value about ten times the number of inks can be adopted. If the ink amount is updated about 10 times the number of inks, it is considered sufficient because the update is performed about 10 times per ink color. Further, since the ink amount is not necessarily updated imax times, imax is set to a sufficiently large number with respect to the number of inks. For example, the number of times about 100 times icmax can be adopted. If it is not determined in step S236 that the variable ic is greater than or equal to icmax, the variable i is incremented in step S238, and the processes in and after step S214 are repeated until the maximum value imax is reached. As described above, according to the loop of steps S214 to S238, the ink amount can be updated a sufficient number of times.

  After step S238, it is determined whether or not the variable ic indicating the number of times the ink amount data has been updated is “0” (step S240). In step S240, it is determined that the variable ic is “0”. If this happens, the processing in steps S242 and S244 is skipped. That is, when the variable ic is “0”, the ink amount has not been updated. Therefore, the system has reached a stable state, and the ink amount is not updated even if the temperature variable is further reduced. Judge.

  If it is not determined in step S240 that the variable ic is “0”, the temperature variable dt is decreased (step S242). In the present embodiment, a constant γ (0 <γ <1) is determined in advance as the smoothing parameter 15e, and dt is updated by a value obtained by multiplying γ by γ. Note that γ is not particularly limited. However, if γ is large, the temperature is lowered slowly, and it is easy to reach a stable state (convergence to an appropriate solution), but the processing time becomes long. Therefore, the size may be determined as appropriate in consideration of the allowable processing time. Actually, for example, a value of about “0.9” can be adopted. Of course, it is only necessary that the temperature can be lowered, and various other methods such as subtracting a constant from dt can be adopted.

  After the temperature is lowered by multiplying the temperature variable dt by a constant γ, the variable d indicating the number of changes of the temperature variable is incremented, and the processing from step S210 onward is repeated until the above-mentioned maximum value dmax is reached (step S244). . If it is determined in step S244 that the variable d has reached the maximum value dmax, and if it is determined in step S240 that the variable ic is “0”, the variable t indicating the number of times the ink amount data is selected. Is incremented to determine whether or not the variable t has reached the maximum value tmax (step S246).

If it is not determined in step S246 that the variable t has reached the maximum value tmax, the processing from step S202 is repeated, and if it is determined in step S246 that the variable t has reached the maximum value tmax, FIG. The process shown in FIG. 3 is terminated, and the process returns to the flow shown in FIG. That is, by changing the amount of ink on ink amount data extracted at random until the maximum value tmax of the selected number, carries out a process to minimize the evaluation function E _s.

  In the ink amount obtained after the above processing, the change in the change rate of the ink amount is smooth. Therefore, when color conversion is performed with reference to this ink amount, a tone jump hardly occurs in the printed matter. . In particular, when a gray gradation is printed, a high-quality printed matter having a smooth gradation change can be obtained. In addition, since the processes of steps S100 to S145 are performed before the smoothing process, when color conversion is performed with reference to the ink amount before the smoothing process, the difference from the target color is very large. Small prints can be obtained.

  Even if the ink amount is changed in the smoothing process, since the evaluation function is defined so as to maintain the original ink amount as much as possible, the LUT 15a is updated with the ink amount data obtained in the above process. For example, it is possible to simultaneously prevent occurrence of tone jump and output in a color close to the target. As a result, it is possible to generate the LUT 15a that can perform printing with very high image quality.

(5) Other Embodiments The LUT 15a created as described above can be used in a printing process that is performed on a general-purpose computer. Accordingly, it can be said that a print control apparatus that performs print processing using the LUT 15a also uses the present invention. FIG. 10 is a block diagram illustrating a configuration example of a computer that uses the LUT 15a during printing. The computer 110 is a general-purpose personal computer, and a PRTDRV 210, an input device driver (DRV) 220, and a display driver (DRV) 230 are incorporated in the OS 200. The display DRV 230 is a driver that controls display of image data and the like on the display 180, and the input device DRV 220 receives code signals from the keyboard 310 and the mouse 320 input via the serial communication I / O 190a, and receives a predetermined signal. A driver that accepts input operations.

  The APL 250 is an application program that can execute retouching of a color image, and the user can operate the input device for operation under the execution of the APL 250 and cause the printing apparatus 400 to print the color image. The printing apparatus 400 is the same type of printing apparatus as the printing apparatus 12 described above. When printing such a color image, the LUT 15a created by the present invention is referred to. The image data 150a of the color image created by the APL 250 is dot matrix data in which each color component of RGB is expressed in gradation, is data compliant with the sRGB standard, and is stored in the HDD 150.

  The PRTDRV 210 is the same software as the PRTDRV 21, and includes an image data acquisition module 210a, a color conversion module 210b, a halftone processing module 210c, and a print data generation module 210d. Further, the LUT 15 a created according to the present invention is stored in the HDD 150. When the user issues a print execution instruction when executing the APL 250, the image data 150a to be printed is acquired by the image data acquisition module 210a, and the image data acquisition module 210a activates the color conversion module 210b. The color conversion module 210b is a module that converts sRGB gradation values into CMYKlclm gradation values, performs an interpolation operation with reference to reference points defined in the LUT 15a, and converts arbitrary sRGB data into CMYKlclm data.

  When the color conversion module 210b performs color conversion to generate CMYKlclm data, the CMYKlclm data is transferred to the halftone processing module 210c. The halftone processing module 210c is a module that performs halftone processing for converting the CMYKlclm tone value of each dot and expressing it with the recording density of the ink droplets, and is a head for attaching ink at the recording density after conversion. Generate drive data. The print data generation module 210d receives the head drive data and performs a rearrangement process in which the print data generation module 210d is rearranged in the order used by the printing apparatus 400. After the rearrangement process, print data is generated by adding predetermined information such as image resolution, and is output to the printing apparatus 400 via the USB I / O 190b. The printing apparatus 400 prints the image displayed on the display 180 based on the print data.

  In this printing process, color conversion is performed with reference to the LUT 15a created according to the present invention, so that it is possible to perform color conversion with a very high accuracy for a specific color, in the case of the above embodiment, gray and surrounding colors. In addition, it is possible to output a printed matter in which gray gradation changes very smoothly. The above description is a very general-purpose printing process by the PRTDRV 210. Therefore, if the LUT 15a according to the present invention is created, the LUT 15a according to the present invention can be replaced with the LUT 15a according to the present invention without changing the hardware configuration at all. Image quality printing can be performed.

  Moreover, embodiment described above is an example for implement | achieving this invention, The structure is not restricted to the said structure. For example, in the above-described correspondence data 15c, CMYKlclm data and its colorimetric values need only be associated with each other with respect to a larger number of colors than the number of colors measured when creating the LUT 15a. Various methods can be adopted as the method.

That is, as shown in FIG. 4C, the present invention is not limited to a configuration in which lattice points at a certain distance from the straight line KW are extracted around the straight line KW. If the ideal value shown in the target data 15b is closer to the other axis than the straight line KW, a lattice point at a certain distance from the other axis is extracted around the other axis substantially parallel to the straight line KW. It is also possible to extract lattice points not in the RGB space but in the L ^* a ^* b ^* space. It is not essential that the grid points to be extracted are evenly arranged in a predetermined color space. For example, a greater number of grid points may be extracted than other colors around the color for which color conversion is to be performed with high accuracy.

Further, the LUT 15a defines the correspondence between the sRGB data and the CMYKlclm data. However, the LUT to be corrected according to the present invention defines the correspondence between the color data in the predetermined color space and the ink amount data. Any configuration can be adopted. For example, it may be an LUT that defines the correspondence between the coordinate data in the L ^* a ^* b ^* space and the ink amount data. Of course, it is not essential that the ink amount data is CMYKlclm data. If the printing apparatus is capable of mounting other ink, the present invention is applied to an LUT that defines ink amount data for specifying the ink amount. It is possible.

Further, as an algorithm to minimize the evaluation function E _s while changing the amount of ink in the smoothing process described above is not limited to the annealing method. For example, an optimization algorithm using a neural network or a genetic algorithm can be employed. That is, if the gradation change in each ink amount data can be evaluated by the evaluation function as in the present invention, the gradation can be smoothed regardless of which algorithm is used for optimization.

It is explanatory drawing which illustrates the correction process of a color conversion table roughly. It is a block diagram of a color conversion table correction apparatus. It is a flowchart of a color conversion table correction process. It is explanatory drawing explaining the example of a process at the time of specifying patch data. It is explanatory drawing explaining a mode that the amount of ink is correct | amended. It is explanatory drawing of the process in a smoothing process part. It is explanatory drawing explaining the view of evaluation by a secondary differentiation. It is explanatory drawing explaining the alternative of an ink. It is a flowchart of a smoothing process. It is a block diagram of a printing control apparatus.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 ... Computer, 11 ... Display, 12 ... Printing apparatus, 15a ... LUT, 15b ... Target data, 15c ... Correspondence relation data, 15d ... Gray data, 15e ... Smoothing parameter, 21 ... PRTDRV, 30 ... LUT correction program, 31 ... Patch data generation unit, 32 ... colorimetric value acquisition unit, 33 ... ink amount data correction unit, 33a ... difference vector calculation unit, 33b ... correction vector calculation unit, 33c ... interpolation calculation unit, 34 ... smoothing processing unit, 40 ... measurement Color machine

Claims (10)

A specific ink amount defined in the color conversion table with reference to a color conversion table that defines the correspondence between the color data in the predetermined color space and the ink amount data indicating the ink amount of each color used in the printing apparatus A color conversion table correction device for correcting data,
Color conversion table data acquisition means for acquiring color conversion table data indicating the color conversion table;
Target data acquisition means for acquiring target data indicating ideal values of colors printed with the specific ink amount data;
Patch printing means for printing a patch by the printing apparatus using the specific ink amount data;
A colorimetric value acquisition means for acquiring a colorimetric value obtained by measuring the printed patch;
In a predetermined device-independent color space, a correction vector is calculated based on a vector having a reverse component of a difference vector indicating a difference between the ideal value and the colorimetric value, and the difference is reduced by the correction vector. An apparatus for correcting a color conversion table, comprising: ink amount data correcting means for calculating ink amount data and correcting the specific ink amount data with the calculated ink amount data. The patch printing unit prints a patch based on the ink amount data corrected by the ink amount data correcting unit, and the colorimetric value acquisition unit acquires a colorimetric value of the patch printed based on the corrected ink amount data, The ink amount data correction unit performs re-correction based on a difference vector indicating a difference between the obtained colorimetric value and the value of the device-independent color space indicating the color corrected by the ink amount data correction unit, The color conversion table correction apparatus according to claim 1, wherein the re-correction is repeatedly performed. 3. The ink amount data correcting means, when performing the re-correction, calculates the correction vector with reference to a difference vector in a correction before the re-correction, and performs the correction. The color conversion table correction device described in 1. The ink amount data correction means determines the correction vector by multiplying the inverse vector of the difference vector by a larger weight as the difference between the colorimetric value and the ideal value is smaller. 4. The color conversion table correction device according to claim 2 or claim 3. Correspondence data acquisition means for acquiring correspondence data defining the correspondence between device-independent color data in a predetermined device-independent color space and the ink amount data,
The ink amount data correcting unit determines a corrected color by the correction vector in the device-independent color space, and calculates ink amount data corresponding to the corrected color with reference to the correspondence data. The color conversion table correction device according to claim 1, wherein the color conversion table correction device is a color conversion table correction device. 6. The color conversion table correction according to claim 5, wherein the correspondence data defines the correspondence for a larger number of colors than the number of colors measured when the color conversion table is created. apparatus. The color according to any one of claims 1 to 6, wherein the target data indicates an ideal value of ink amount data for printing a plurality of colors which are achromatic and have different brightness. Conversion table correction device. A specific ink amount defined in the color conversion table with reference to a color conversion table that defines the correspondence between the color data in the predetermined color space and the ink amount data indicating the ink amount of each color used in the printing apparatus A color conversion table correction method for correcting data,
A color conversion table data acquisition step for acquiring color conversion table data indicating the color conversion table;
A target data acquisition step of acquiring target data indicating an ideal value of a color printed with the specific ink amount data;
A patch printing step of printing a patch by the printing apparatus using the specific ink amount data;
A colorimetric value acquisition step for acquiring a colorimetric value obtained by measuring the color of the printed patch;
In a predetermined device-independent color space, a correction vector is calculated based on a vector having a reverse component of a difference vector indicating a difference between the ideal value and the colorimetric value, and the difference is reduced by the correction vector. A color conversion table correction method comprising: an ink amount data correction step of calculating ink amount data and correcting the specific ink amount data with the calculated ink amount data. A specific ink amount defined in the color conversion table with reference to a color conversion table that defines the correspondence between the color data in the predetermined color space and the ink amount data indicating the ink amount of each color used in the printing apparatus A color conversion table correction program for correcting data,
A color conversion table data acquisition function for acquiring color conversion table data indicating the color conversion table;
A target data acquisition function for acquiring target data indicating an ideal value of a color printed with the specific ink amount data;
A patch printing function for causing the printing device to print a patch based on the specific ink amount data;
A colorimetric value acquisition function for acquiring colorimetric values obtained by measuring colors of the printed patches;
In a predetermined device-independent color space, a correction vector is calculated based on a vector having a reverse component of a difference vector indicating a difference between the ideal value and the colorimetric value, and the difference is reduced by the correction vector. A color conversion table correction program characterized in that a computer realizes an ink amount data correction function that calculates ink amount data and corrects the specific ink amount data based on the calculated ink amount data. A print control device that converts color data in a predetermined color space into ink amount data indicating the amount of ink of each color used in the printing device, and causes the printing device to perform printing based on the ink amount data,
Image data acquisition means for acquiring image data expressing the color of each pixel by color data in the predetermined color space;
A color conversion table defining a correspondence relationship between the color data in the predetermined color space and the ink amount data indicating the ink amount of each color used in the printing apparatus, and the specific ink amount data In an independent color space, ink amount data for reducing the difference is calculated by a correction vector calculated based on a vector having a reverse component of the difference vector indicating the difference between the ideal value and the colorimetric value. A color conversion means for converting the color of each pixel indicated by the image data into ink amount data with reference to the color conversion table corrected by
A printing control apparatus comprising: a printing execution unit that executes printing based on the converted ink amount data.

Images