Color prediction model creating method, color prediction model creating device, printing system, and computer program

Abstract

To provide a technique for creating a color prediction model taking into consideration the priorities of work time and prediction accuracy.SOLUTION: A method for creating a color prediction model includes a step of (a) acquiring color chart data representing N color patch groups, a step of (b) (i) creating n sets of provisional learning data sets corresponding to the n color patch groups using predicted spectral reflectance predicted by a simple color prediction model, (ii) creating a provisional color prediction model by learning using n sets of provisional learning data sets, and (iii) successively executing accuracy calculation processing for calculating the prediction accuracy of the provisional color prediction model while increasing an integer n, a step of (c) determining the number nc of color patch groups suitable for learning of the color prediction model from the relationship between the integer n and the prediction accuracy, a step of (d) performing printing and measuring the spectral reflectance for nc color patch groups, and creating learning data including the measurement results of the spectral reflectance, and a step of (e) performing learning of the color prediction model.SELECTED DRAWING: Figure 3

Description

[Technical field]

[0001] The present disclosure relates to a method for creating a color prediction model, a color prediction model creating device, a printing system, and a computer program. [Background technology]

[0002] In the printing device, a color conversion process is performed to convert the color values ​​of the input image data expressed in a first color system into color values ​​in a second color system corresponding to the type of ink. In the color conversion process, a plurality of color conversion tables for associating the color values ​​of the first color system with the color values ​​of the second color system are referenced. When creating such a color conversion table, a color prediction model is used that predicts the spectral reflectance of a printed matter printed with an arbitrary ink amount set from the ink amount set. Patent Document 1 disclosed by the applicant of the present disclosure discloses a method for creating a teacher data set for a color prediction model by repeating data selection and learning from a group of virtual teacher data candidates created using a pre-model. According to this method, the prediction accuracy of the color prediction model can be efficiently improved by additionally selecting teacher data for the color prediction model. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2023-128280 A Summary of the Invention [Problem to be solved by the invention]

[0004] In recent years, it is not always desirable to spend a long time working to create a color prediction model with the highest accuracy, and there are cases where it is desirable to create a color prediction model in a short amount of time even if the prediction accuracy is somewhat inferior. Also, as the use of color prediction models becomes more widespread, it has become necessary to balance prediction accuracy and work time depending on the application. [Means for solving the problem]

[0005] According to a first aspect of the present disclosure, there is provided a method for creating a color prediction model for predicting spectral reflectance from an ink amount set of a plurality of types of ink constituting an ink set, the method comprising the steps of: (a) acquiring color chart data representing N color patch groups, where N is an integer equal to or greater than 2; (b) (i) creating n sets of provisional learning data sets corresponding to the n color patch groups using predicted spectral reflectances predicted by a simplified color prediction model that can be trained more easily than the color prediction model, (ii) creating a provisional color prediction model by training the color prediction model using the n sets of provisional learning data sets, and (iii) calculating the prediction accuracy of the provisional color prediction model, while sequentially increasing the integer n; and (c) calculating the number n of the color patch groups suitable for training the color prediction model from the relationship between the integer n and the prediction accuracy. c (d) determining said n c The method includes the steps of (a) printing the color patch groups using a printing device and measuring the spectral reflectance, and creating learning data including the results of the spectral reflectance measurement; and (b) using the learning data to perform learning of the color prediction model.

[0006] According to a second aspect of the present disclosure, there is provided a color prediction model creation device that creates a color prediction model that predicts spectral reflectance from an ink amount set of multiple types of inks that constitute an ink set. The color prediction model creation device includes a data acquisition unit that acquires color chart data representing N color patch groups, where N is an integer equal to or greater than 2, an accuracy calculation unit that sequentially executes an accuracy calculation process while increasing the integer n, the accuracy calculation unit performing the following steps: (i) creating n sets of temporary learning data sets corresponding to the n color patch groups using predicted spectral reflectances predicted by a simplified color prediction model that can be learned more easily than the color prediction model, (ii) creating a temporary color prediction model by learning the color prediction model using the n sets of temporary learning data sets, and (iii) calculating the prediction accuracy of the temporary color prediction model; and (ii) calculating the number n of the color patch groups suitable for learning the color prediction model from the relationship between the integer n and the prediction accuracy. c A determination unit that determines c The system includes a learning data creation unit that performs printing using a printing device and measuring the spectral reflectance of the color patch groups, and creates learning data including the measurement results of the spectral reflectance, and a learning unit that uses the learning data to learn the color prediction model.

[0007] According to a third aspect of the present disclosure, a printing system is provided. The printing system includes a color prediction model creation device that creates a color prediction model that predicts spectral reflectance from an ink amount set of multiple types of ink that constitute an ink set, and a printing device. The color prediction model creation device includes a data acquisition unit that acquires color chart data representing N color patch groups, where N is an integer equal to or greater than 2, an accuracy calculation unit that sequentially executes an accuracy calculation process while increasing the integer n, the accuracy calculation unit performing the following steps, where n is an integer equal to or greater than 1 and equal to or less than N: (i) creating n sets of temporary learning data sets corresponding to the n color patch groups using predicted spectral reflectances predicted by a simplified color prediction model that can be learned more easily than the color prediction model, (ii) creating a temporary color prediction model by learning the color prediction model using the n sets of temporary learning data sets, and (iii) calculating the prediction accuracy of the temporary color prediction model; and (ii) calculating the number n of the color patch groups suitable for learning the color prediction model from the relationship between the integer n and the prediction accuracy. c A determination unit that determines c The color prediction model includes a learning data creation unit that performs printing by the printing device and measurement of spectral reflectance for each of the color patch groups, and creates learning data including the measurement results of the spectral reflectance, and a learning unit that uses the learning data to learn the color prediction model.

[0008] According to a fourth aspect of the present disclosure, there is provided a computer program for creating a color prediction model that predicts spectral reflectance from an ink amount set of a plurality of types of ink that constitute an ink set. This computer program includes: (a) a process of acquiring color chart data representing N color patch groups, where N is an integer equal to or greater than 2; (b) a process of sequentially executing, while increasing the integer n, a calculation process of calculating the number n of color patch groups suitable for training the color prediction model, where n is an integer equal to or greater than 1 and equal to or less than N, (i) creating n sets of provisional training data sets corresponding to the n color patch groups using predicted spectral reflectances predicted by a simplified color prediction model that can be trained more easily than the color prediction model, (ii) training the color prediction model using the n sets of provisional training data sets, and (iii) calculating the prediction accuracy of the provisional color prediction model, and (c) calculating, from the relationship between the integer n and the prediction accuracy, the number n of the color patch groups suitable for training the color prediction model. c (d) determining said n c The computer is caused to execute a process of (a) printing the color patch groups using a printing device, measuring the spectral reflectance, and creating learning data including the results of the measurement of the spectral reflectance, and (b) learning the color prediction model using the learning data. [Brief description of the drawings]

[0009] [Figure 1] FIG. 1 is a block diagram of a printing system. [Diagram 2] FIG. 4 is an explanatory diagram showing an example of color chart information. [Diagram 3] 5 is a flowchart showing the procedure of a process for creating a color prediction model in the first embodiment. [Figure 4] FIG. 4 is an explanatory diagram showing an example of a print setting window. [Diagram 5] FIG. 4 is an explanatory diagram showing an example of a color measurement setting window. [Figure 6] FIG. 11 is an explanatory diagram showing an example of a window for selecting color chart data. [Figure 7] FIG. 4 is an explanatory diagram showing the relationship between the number of color charts and prediction accuracy in the first embodiment. [Figure 8] 11 is an explanatory diagram showing an example of changing the number of times of learning a temporary color prediction model depending on the number of color charts. [Figure 9] 11 is an explanatory diagram showing an example of changing the number of times of learning a temporary color prediction model depending on prediction accuracy. [Figure 10] FIG. 1 is an explanatory diagram showing an example of a printing device including a color measurement device. [Figure 11] 10 is a flowchart showing the procedure of a process for creating a color prediction model in a second embodiment. [Figure 12] FIG. 13 is an explanatory diagram showing an example of a window for specifying the number of color charts in the second embodiment. [Figure 13] 13 is a flowchart showing the procedure of a color prediction model creation process in a third embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] A. First embodiment: 1 is a block diagram showing a schematic configuration of a printing system according to an embodiment of the present invention. The printing system 500 includes a color prediction model creation device 100, a printing device 200, a color measurement device 300, and a display device 400.

[0011] The color prediction model creation device 100 creates a color prediction model CM that predicts the spectral reflectance of a printed material printed with an ink amount set used in the printing device 200, from that ink amount set. An "ink amount set" refers to a combination of ink amounts of multiple types of ink that can be used in the printing device 200. A combination of multiple types of ink that can be used in the printing device 200 is called an "ink set."

[0012] The color prediction model creation device 100 is a computer including a CPU 50, a storage unit 60, and an input / output interface 70. The CPU 50, the storage unit 60, and the input / output interface 70 are connected via an internal bus so as to be able to communicate bidirectionally.

[0013] The CPU 50 executes a color prediction model creation program 61 pre-stored in the storage unit 60, thereby functioning as a processing setting unit 51, a data acquisition unit 52, an accuracy calculation unit 53, a determination unit 54, a learning data creation unit 55, and a learning unit 56. At least a part of the functions of these units 51 to 56 may be realized by a hardware circuit, or may be realized on the cloud.

[0014] The process setting unit 51 receives various process settings related to the process of creating the color prediction model CM. The contents of the process settings will be described later.

[0015] The data acquisition unit 52 acquires one or more color chart data 62 selected from the plurality of color chart data 62. When N is an integer equal to or greater than 2, one color chart data 62 is data representing N color patch groups. In this embodiment, it is assumed that one color patch group corresponds to one color chart CC. However, one color patch group may be defined in a unit different from one color chart CC. For example, a collection of a predetermined number of color patches may be defined as a "color patch group."

[0016] The accuracy calculation unit 53 sequentially executes an accuracy calculation process including the following steps (i) to (iii), where n is an integer between 1 and N, while increasing the integer n. (i) Using predicted spectral reflectances predicted by a simplified color prediction model SM that can be trained more easily than the color prediction model CM, n sets of provisional training data sets corresponding to n groups of color patches are created. (ii) A temporary color prediction model PM is created by training the color prediction model CM using n sets of temporary training data sets. (iii) The prediction accuracy of the temporary color prediction model PM is calculated. The accuracy calculation process will be described in detail later.

[0017] The meaning of the simplified color prediction model SM being "more easily trainable than the color prediction model CM" is that the amount of training data is smaller than that of the color prediction model CM, and the training time is shorter. The simplified color prediction model SM is a model that predicts the spectral reflectance from an ink amount set, and can be configured as a color prediction model based on a known spectral Neugebauer model, for example. As a color prediction model based on a spectral Neugebauer model, the printing model described in JP-A-2006-334945 can be used. Note that a color prediction model other than the spectral Neugebauer model may be used as the simplified color prediction model SM. However, it is preferable that the simplified color prediction model SM requires a smaller amount of training data and a shorter training time than the finally created color prediction model CM. It is preferable that the simplified color prediction model SM has been trained using training data for the simplified color prediction model SM that has been created in advance.

[0018] The determination unit 54 determines the number n of color patch groups suitable for learning the color prediction model CM from the relationship between the integer n and the prediction accuracy. c Determine the number n c "Optimal number n c As described above, in this embodiment, one color patch group corresponds to one color chart CC, so the number of color patch groups n c corresponds to the number of color charts CC.

[0019] The learning data creation unit 55 is c For each color patch group, printing is performed by the printing device 200 and the spectral reflectance R(λ) is measured by the color measurement device 300 to generate learning data including the measurement results of the spectral reflectance R(λ). c The data indicates the correspondence relationship between the ink amount set and the spectral reflectance R(λ) for each color patch of the color patch group. The learning unit 56 uses the learning data to execute learning of the color prediction model CM.

[0020] In addition to the color prediction model creation program 61, the storage unit 60 stores a plurality of color chart data 62, color chart information 63, and accuracy confirmation data 64. The color chart data 62 is image data representing a color chart including a plurality of single-color color patches and a plurality of mixed-color color patches. The pixel values ​​of the color chart data 62 are represented by ink amount sets. In this embodiment, a plurality of color chart data 62 are created in advance and stored in the storage unit 60. The color chart information 63 is information relating to each individual color chart data 62. The contents of the color chart information 63 will be described later.

[0021] The accuracy confirmation data 64 is data used when calculating the prediction accuracy of the temporary color prediction model PM. The accuracy confirmation data 64 is data that indicates the correspondence between ink amount sets and measured values ​​of spectral reflectance for a plurality of ink amount sets. The spectral reflectance included in the accuracy confirmation data 64 is preferably a measured value for a color patch printed on the same printing medium as the printing medium to which the color prediction model CM is applied.

[0022] The printing device 200 has a plurality of ink containers 210, each of which contains one type of ink. In this embodiment, the number of ink containers 210 is six, and an ink set containing six types of ink, CMYKLcLm, can be used. Here, C is cyan, M is magenta, Y is yellow, K is black, Lc is light cyan, and Lm is light magenta process ink. The printing device 200 may be configured to be able to use spot color inks. Also, not only the ink color type but also the number of ink colors can be set arbitrarily.

[0023] The input / output interface 70 transmits images such as the color chart data 62 to the printing device 200, and also receives color measurement results of the print image printed on the printing medium P from the color measurement device 300. The input / output interface 70 further transmits display images to the display device 400, and displays various windows and processing results, which will be described later.

[0024] The printing device 200 is an inkjet printer that prints an image on the printing medium P by ejecting ink onto the printing medium P. Upon receiving color chart data 62 from the color prediction model creation device 100, the printing device 200 prints a color chart CC on the printing medium P in response to the color chart data 62.

[0025] The color measurement device 300 measures the color of a printed matter produced by the printing device 200. The color measurement device 300 measures the spectral reflectance R(λ) of each of a plurality of color patches included in the color chart CC. The measured spectral reflectance R(λ) is supplied to the color prediction model creation device 100.

[0026] 2 is an explanatory diagram showing an example of the color chart information 63. The color chart information 63 is information in which five items are registered for each color chart data 62: (1) media name, (2) media surface type, (3) ink set, (4) color chart data name, and (5) auxiliary information.

[0027] In this disclosure, "media" refers to print media. In the example of Fig. 2, four media names, Photo_A, Photo_B, Mat_A, and Mat_B, are presented. As described in the media surface type column, Photo_A and Photo_B are glossy paper, and Mat_A and Mat_B are matte paper. Two types of ink sets, CMYK and CMYKLcLm, are presented.

[0028] "Ink set CMYK" means one of the following: (a) Ink set for a printing device equipped with only CMYK inks. (b) An ink set used in a printing mode that does not use inks other than CMYK (Lc ink and Lm ink) among the CMYKLcLm inks installed in the printing device. It is usually used in the sense of (a) above, but it may also be used in the sense of (b) above.

[0029] As auxiliary information, whether or not the color chart data for Photo_A and Photo_B can be used to create a color prediction model for Photo_C is registered. Specifically, the color chart data for Photo_A can be used to create a color prediction model for Photo_C, and the color chart data for Photo_B cannot be used to create a color prediction model for Photo_C. Photo_C is a medium that has color development characteristics similar to those of Photo_A.

[0030] The reason why color chart data for Photo_C is not prepared and the fact that it can be substituted with other media is registered in the auxiliary information is that it is practically difficult to prepare color chart data for all media that are expected to be used. For this reason, in this embodiment, it is assumed that color chart data will be prepared for representative media and that the color chart data will be used for media with similar characteristics.

[0031] It should be noted that some of the five items illustrated in FIG. 2 may be omitted, for example, the auxiliary information may be omitted.

[0032] In the first embodiment, one piece of color chart data 62 is selected using the color chart information 63. However, it is also possible to automatically select color chart data to be referenced when print settings are made, which will be described later, without using the color chart information 63.

[0033] 3 is a flowchart showing the procedure of the process of creating a color prediction model. In step S10, the process setting unit 51 accepts various process settings related to the process of creating a color prediction model CM. In the first embodiment, the process settings include (1) print settings, (2) color measurement settings, and (3) selection of color chart data. These process settings will be described in order below.

[0034] 4 is an explanatory diagram showing an example of a print setting window W1. Using this window W1, the user can set four items related to printing the color chart CC: (1) media name, (2) number of passes, (3) duty limit value, and (4) whether or not to perform color correction.

[0035] "Number of passes" refers to the number of main scans of the print head required to print all the pixels on one main scan line. For example, "4 passes" means that all the pixels on one main scan line are printed by performing four main scans. It is also possible to select, for example, 1 pass or 2 passes as the number of passes.

[0036] The "duty limit value" is the upper limit of the total amount of ink permitted to be ejected per pixel. In the example of Figure 4, the duty limit value to be used for actual printing can be specified, taking into consideration cases where the media selected in the print settings differs from the media used for actual printing. In this case, the amount of ink ejected per pixel during actual printing is corrected as follows, taking into consideration the specified duty limit value. Vc = V × (Dt_set / Dt_pre) …(q1) Here, Vc is the corrected ink amount, V is the original ink amount, Dt_set is the duty limit value specified in the print settings, and Dt_pre is the duty limit value set when the color chart data 62 was created. Normally, Dt_set is set to a value equal to or less than Dt_pre. However, Dt_set may be set to a value greater than Dt_pre.

[0037] For "color correction," you can specify either no color correction or color correction. The "no color correction" setting means that when using a RIP (Raster Image Processor) or other process to print a color chart CC, the color management function is turned off so that no color correction is performed.

[0038] Note that some of the four items that can be set in the print setting window W1 may be omitted, and other items may be added. This also applies to the windows for other process settings described below.

[0039] 5 is an explanatory diagram showing an example of the color measurement setting window W2. Using this window W2, the user can set three items related to the spectral reflectance measurement of the printed color chart CC: (1) the measurement device, (2) the measurement method, and (3) the number of measurements.

[0040] "Measurement device" is a setting of identification information of the color measurement device 300. As the identification information, a model number or a device name can be used.

[0041] As the "measurement method," a measurement method that can be performed by the colorimetric device 300 can be selected. Two measurement methods are generally known: spot colorimetry and scan colorimetry. Spot colorimetry is a method in which the colorimetric device stops at each color patch to measure the color. Scan colorimetry is a method in which multiple color patches are read continuously while the measurement device is moved. Scan colorimetry can complete the measurement of one page's worth of color patches in significantly less work time than spot colorimetry.

[0042] The "number of measurements" can be set to one, or any number of measurements greater than or equal to two. In general, the more measurements are made, the less variation there is in the measurement results of the spectral reflectance, and the more accurate the predictions of the color prediction model CM will be, but the more time it will take to complete the work. Normally, the color measurement device itself is a measuring instrument that has measurement variations, so it is preferable to perform measurements multiple times and calculate the statistical values ​​of the measurements to suppress the measurement variations. When the number of measurements is set to multiple times, in the case of spot colorimetry, the same color is measured a specified number of times before moving to the next color, but in the case of scan colorimetry, the row is scanned a specified number of times. This is also a factor that affects the work time.

[0043] FIG. 6 is an explanatory diagram showing an example of a window W3 for selecting color chart data 62. When the user selects an ink set in this window W3, the attributes of multiple color chart data suitable for that ink set are displayed. In this example, the media name, the surface type of the media, and auxiliary information are displayed as attributes of each color chart data. These attributes are information registered in the color chart information 63 shown in FIG. 2. Also, the window W3 is provided with check boxes CB. The user can select one color chart data by checking one check box CB.

[0044] In the print settings in Fig. 4 described above, Photo_C is selected as the medium, and the plurality of color chart data 62 in Fig. 2 does not include any data with Photo_C as the medium. Even in this case, auxiliary information indicating that it can be used for Photo_C is shown in Fig. 2 and Fig. 6, so the user can select one piece of color chart data suitable for Photo_C.

[0045] When Photo_C is selected as the medium in the print settings in Fig. 4, the window W3 in Fig. 6 may be displayed in a state where color chart data that can use Photo_C is automatically selected. In this way, it is possible to automatically present candidates for color chart data that are suitable for the medium selected in the print settings.

[0046] As a list for selecting color chart data 62, a list as shown in Fig. 2 may be used instead of a list as shown in Fig. 6. All of these lists have in common that they include, as setting information suitable for each color chart data 62, print medium information indicating one print medium among a plurality of print media usable in the printing device 200, and ink set information indicating one ink set among a plurality of ink sets usable in the printing device 200. By using such a list, it is possible to select color chart data 62 suitable for a combination of print medium and ink set as the color chart data 62 to be used in creating the color prediction model CM.

[0047] Among the various processing settings described with reference to FIGS. 4 to 6, items other than the selection of color chart data may be omitted as appropriate.

[0048] When the above various settings are made by the user, the process proceeds to step S20 in Fig. 3, where the data acquisition unit 52 acquires the selected color chart data 62 from the storage unit 60. As described above, one piece of color chart data 62 is data representing N sheets of color charts CC.

[0049] Steps S31 to S35 between steps S30s and S30e are a routine that repeats processing while increasing a parameter n, which indicates the number of color charts, by a predetermined increment in the range of 1 to N. The increment can be set to any integer equal to or greater than 1. This routine corresponds to the accuracy calculation processing executed by accuracy calculation unit 53. Note that the initial value of parameter n may be set to a value greater than 1. In the following description, the "number of color charts n" will also be referred to simply as the "number of charts n".

[0050] In step S31, the accuracy calculation unit 53 calculates predicted values ​​of the spectral reflectance for the n color charts CC using the simplified color prediction model SM. This predicted value is called the "predicted spectral reflectance." Each color chart CC includes multiple color patches, and the predicted spectral reflectance is obtained for each of these color patches.

[0051] In step S32, the accuracy calculation unit 53 uses the predicted spectral reflectance to create n sets of temporary learning data sets corresponding to n color charts CC. One set of temporary learning data sets is a collection of temporary learning data related to multiple color patches included in one color chart CC. One piece of temporary learning data is data indicating the relationship between the ink amount set and the predicted spectral reflectance for one color patch. The reason why "temporary" is added before "learning data" is that the spectral reflectance is not a measured value but a predicted value. In step S33, the accuracy calculation unit 53 executes learning of the temporary color prediction model PM using the n sets of temporary learning data sets.

[0052] In step S34, the accuracy calculation unit 53 calculates the prediction accuracy of the provisional color prediction model PM using the accuracy confirmation data 64. As an index of prediction accuracy, for example, the average value of the color difference between the first L*a*b* value calculated from the predicted value of the spectral reflectance and the second L*a*b* value calculated from the actual measured value of the spectral reflectance can be used. The predicted value of the spectral reflectance can be obtained by inputting the ink amount set of the accuracy confirmation data 64 into the trained provisional color prediction model PM. The measured value of the spectral reflectance can be the value included in the accuracy confirmation data 64. The reason why the "average value of the color difference" is obtained is that the accuracy confirmation data 64 indicates the relationship between the ink amount set and the measured value of the spectral reflectance for a plurality of color patches, and therefore the color difference is calculated for these plurality of color patches. The color difference ΔE as an index of prediction accuracy is also called the "predicted color difference". The higher the prediction accuracy, the smaller the predicted color difference.

[0053] In step S35, the accuracy calculation unit 53 judges whether the prediction accuracy of the temporary color prediction model PM has reached the target accuracy. If the prediction accuracy has not reached the target accuracy, the process returns from step S30e to step S30s, where the accuracy calculation unit 53 increments the number of color charts n by one and executes the processes from step S31 onwards again. On the other hand, if the prediction accuracy has reached the target accuracy, the process proceeds to step S36, where the accuracy calculation unit 53 increments the current number of color charts n by one, equal to the optimal number n. c It is determined as follows.

[0054] 7 is an explanatory diagram showing the relationship between the number of color charts and prediction accuracy in the first embodiment. Here, the predicted color difference ΔEp is used as an index representing the prediction accuracy. In the example of FIG. 7, when the number of color charts n is set to 3, the predicted color difference ΔEp is equal to or smaller than the target color difference ΔEt, and the prediction accuracy reaches the target accuracy. Therefore, the optimal number n c is determined to be 3.

[0055] As described above, if the prediction accuracy does not reach the target accuracy in step S35, the process returns from step S30e to step S30s, and the accuracy calculation unit 53 increments the number of color charts n by 1 and executes the processes from step S31 onwards again. If the prediction accuracy does not reach the target accuracy until the number of color charts n reaches the maximum value N, the process proceeds to step S40. In step S40, the accuracy calculation unit 53 calculates the optimal number n of color charts that provides the best prediction accuracy among the number n of color charts equal to or less than N. c Determine.

[0056] In step S36 or S40, the optimal number n c When n is determined, the process proceeds to step S50, in which the learning data creation unit 55 calculates n c For one color chart, printing is performed by the printing device 200 and color measurement is performed by the color measurement device 300 to create learning data including the measurement results of the spectral reflectance R(λ). This learning data is also called "actual learning data." Each data item of the actual learning data includes an ink amount set and a measurement value of the spectral reflectance R(λ). The measurement of the spectral reflectance R(λ) is performed by the color measurement device 300. Note that the color measurement device 300 may be implemented within the printing device 200.

[0057] FIG. 10 is an explanatory diagram showing an example of a printing device 200 including a color measurement device 300. In this example, a plurality of color patches CP of a color chart CC are printed on a printing medium P by a head scanning mechanism 220 of the printing device 200. In the feed direction FD of the printing medium P, the color measurement device 300 is disposed at a position downstream of the head scanning mechanism 220. Therefore, immediately after the color patches CP are printed, it is possible to measure the spectral reflectance R(λ) using the color measurement device 300 in the printing device 200. If the printing device 200 has a configuration capable of connecting to a server via a network, the measured spectral reflectance R(λ) may be transmitted to the server, and a color prediction model creation unit implemented in the server may be used to automatically create a color prediction model CM.

[0058] In step S60, the learning unit 56 uses the actual learning data to learn the color prediction model CM. The color prediction model CM can be, for example, a regression model with multidimensional output configured by a neural network. By executing machine learning using the actual learning data, such a color prediction model CM becomes an inference model that can predict the spectral reflectance from an arbitrary ink amount set of an ink set.

[0059] It should be noted that the learning of the temporary color prediction model PM in step S33 can be performed by changing the number of learning times or the learning method. Here, the "number of learning times" means the number of epochs, that is, "how many times the same learning data is repeatedly learned." The number of learning times may be determined by applying, for example, the following method. (1) The number of learning sessions is set to a constant value. (2) The number of learning times is changed according to the number of color charts n. (3) The number of learning times is changed according to the prediction accuracy (predicted color difference ΔEp) of the temporary color prediction model PM.

[0060] 8 is an explanatory diagram showing an example of changing the number of times the temporary color prediction model PM is trained depending on the number of color charts. In this example, the number of times the temporary color prediction model PM is trained is reduced as the number of color charts n increases. In this way, even if the number of color charts n increases, it is possible to prevent the work time required for training from becoming excessively long.

[0061] 9 is an explanatory diagram showing an example of changing the number of times of learning of the temporary color prediction model PM depending on the prediction accuracy. In this example, learning is terminated when the decrease width δ of the predicted color difference ΔEp of the temporary color prediction model PM when the number of times of learning is increased becomes equal to or smaller than a threshold value δth. In this way, learning that does not contribute to improving the prediction accuracy can be omitted, and the overall work time can be shortened.

[0062] As a learning method, a method of additional learning may be used for the trained temporary color prediction model PM using transfer learning or fine tuning. For example, a temporary learning data set for the (n+1)th color chart may be added to a temporary color prediction model PM that has been trained using n sets of temporary learning data sets for n color charts, and learning may be performed. In this way, the work time required for training the temporary color prediction model PM can be shortened.

[0063] As described above, in the first embodiment, the number of color charts n that can obtain sufficient prediction accuracy using the simplified color prediction model SM is c It is also possible to determine n c Since a sheet of color chart is printed and the color prediction model CM is created using the measurement results of the spectral reflectance, the work time required to create the color prediction model CM can be efficiently shortened.

[0064] B. Second embodiment: Fig. 11 is a flowchart showing the procedure of the process of creating a color prediction model in the second embodiment. The device configuration of the second embodiment is the same as that of the first embodiment. The process of Fig. 11 differs from the process of the first embodiment shown in Fig. 3 in only the following two points. (1) Steps S35 and S36 are omitted. (2) Step S40 is replaced by step S41.

[0065] In the second embodiment, the accuracy calculation process of steps S31 to S34 is continuously executed until the number of color charts n reaches N. Then, in step S41, the optimal number of color charts n c In one example, the accuracy calculation unit 53 displays the relationship between the number of color charts n and the prediction accuracy on the display device 400, and the optimal number n is determined by the user. c Accepts specification of.

[0066] 12 is an explanatory diagram showing an example of a window W4 for specifying the number of color charts in the second embodiment. This window W4 includes a graph showing the relationship between the number of color charts n and the predicted color difference ΔEp, and an optimum number n c The input field IFD is used to input the optimum number n of sheets that can achieve the target color difference ΔEt by inputting an integer into the input field IFD. c You can specify the optimal number of sheets n by using the marker MK in the graph. c In addition, the display of the target color difference ΔEt may be omitted.

[0067] In this way, in the second embodiment, the accuracy calculation unit 53 continues to execute the accuracy calculation process until the number n of color charts reaches N, displays the relationship between the number n of color charts and the predicted accuracy, and determines the optimal number n by the user. c As a result, the user can select the number of color charts n suitable for training the color prediction model CM, taking into account the balance between prediction accuracy and work time. c In other words, it is possible to reflect the user's intention, such as "prioritizing time and reducing the number of sheets" or "prioritizing accuracy and increasing the number of sheets."

[0068] In step S41, the optimal number n c Instead of accepting the designation of n, the accuracy calculation unit 53 calculates the number n of color charts that will provide the best prediction accuracy as the optimal number n cAs in the first embodiment, a plurality of color patches separated by a unit other than a single color chart may be used as one color patch group.

[0069] C. Third embodiment: Fig. 13 is a flowchart showing the procedure of the process of creating a color prediction model in the third embodiment. The device configuration of the third embodiment is the same as that of the first and second embodiments. The process of Fig. 13 differs from the process of the second embodiment shown in Fig. 12 in only the following two points. (1) Before step S20 of acquiring color chart data, a start step S100s of a routine for selecting one by one from all color chart data to be processed is inserted, and a corresponding end step S100e is inserted after step S30e. (2) In the process setting in step S10a, the designation of the medium described with reference to FIG. 4 and the selection of the color chart data described with reference to FIG. 6 are not performed.

[0070] In step S41 of the third embodiment, the accuracy calculation unit 53 calculates the number n of color charts that can achieve the best prediction accuracy for the color chart data 62 that provides the best prediction accuracy as the optimal number n c In this way, the accuracy calculation process does not require user specification, and the optimal number of color charts n c can be determined automatically.

[0071] However, as in the second embodiment, the optimal number n c In addition, similarly to the first embodiment, a plurality of color patches separated by a unit other than a single color chart may be used as one color patch group.

[0072] 2, the multiple color chart data 62 include data corresponding to different types of media. It is preferable to use all of the multiple color chart data 62 prepared in advance as the processing target of the accuracy calculation process in the third embodiment, regardless of the type of media. In this way, it is possible to create learning data for the color prediction model CM using color chart data 62 with good prediction accuracy among all the color chart data 62.

[0073] Note that the prepared color chart data 62 does not necessarily include color chart data for all media. Therefore, when using an unknown medium, if the prepared color chart data 62 is used as the target of the accuracy calculation process, there is an advantage in that a medium having the closest color and characteristics can be selected from a plurality of media suitable for the color chart data 62, and the color chart data 62 can be used.

[0074] However, in the process setting of step S10a, if the user specifies the medium as described in Fig. 4, one or more color chart data 62 suitable for that medium may be selected, and the processes of steps S100s to S100e of Fig. 13 may be executed for those color chart data 62. In this case, too, the user's specification is not required, and the optimal number n of color charts is automatically set. c This has the advantage that it can be determined automatically.

[0075] Other forms: The present disclosure is not limited to the above-mentioned embodiment, and can be realized in various forms without departing from the spirit of the present disclosure. For example, the present disclosure can be realized in the following forms. The technical features in the above-mentioned embodiments corresponding to the technical features in each form described below can be appropriately replaced or combined in order to solve some or all of the problems of the present disclosure, or to achieve some or all of the effects of the present disclosure. Furthermore, if the technical feature is not described as essential in this specification, it can be appropriately deleted.

[0076] (1) According to a first aspect of the present disclosure, there is provided a method for creating a color prediction model for predicting spectral reflectance from an ink amount set of a plurality of types of ink constituting an ink set, the method comprising the steps of: (a) acquiring color chart data representing a group of N color patches, where N is an integer equal to or greater than 2; (b) (i) creating n sets of provisional learning data sets corresponding to the n groups of color patches using predicted spectral reflectances predicted by a simplified color prediction model that can be learned more easily than the color prediction model, (ii) creating a provisional color prediction model by learning the color prediction model using the n sets of provisional learning data sets, and (iii) calculating the prediction accuracy of the provisional color prediction model, while sequentially increasing the integer n; and (c) calculating the number n of the color patch groups suitable for learning the color prediction model from the relationship between the integer n and the prediction accuracy. c (d) determining said n c The method includes the steps of (a) printing the color patch groups using a printing device and measuring the spectral reflectance, and creating learning data including the results of the spectral reflectance measurement; and (b) using the learning data to perform learning of the color prediction model. According to this method, the number of color patch groups n that can achieve sufficient prediction accuracy using the simplified color prediction model is c can be determined, n c Since the color prediction model is created using the printing and spectral reflectance measurement results for each color patch group, the work time required to create the color prediction model can be efficiently reduced.

[0077] (2) In the above method, the step (b) includes a step of stopping the increment of the integer n when the prediction accuracy reaches a target accuracy or less, and the step (c) is a step of changing the integer n when the prediction accuracy reaches a target accuracy or less to the number n of the color patch set. c The method may include a step of determining: According to this method, the number of color patch groups for which the prediction accuracy is equal to or lower than the target accuracy can be easily determined, and the work time can be reduced.

[0078] (3) In the above method, the step (b) is continuously executed until the integer n reaches the integer N, and the step (c) includes: (c1) displaying a relationship between the integer n and the prediction accuracy; and (c2) selecting the number n of the color patch group by a user. c and receiving a designation of the selected item. According to this method, the user can determine the number of color patch groups used for training the color prediction model, taking into consideration the balance between prediction accuracy and work time.

[0079] (4) In the above method, the step (a) includes a step of sequentially selecting one color chart data to be used for creating the color prediction model from a plurality of color chart data created in advance, the step (b) is executed for each of the selected color chart data, and the step (c) includes a step of sequentially selecting the number n of the color patch groups for one color chart data having the best prediction accuracy among the plurality of color chart data. c The method may include determining: According to this method, a color prediction model can be created by using one color chart data having the best prediction accuracy among a plurality of color chart data, while balancing prediction accuracy and work time.

[0080] (5) In the above method, one of the color patch groups may be one color chart printed on one sheet of printing medium. According to this method, a number suitable for creating a color prediction model can be determined for each number of color charts.

[0081] (6) In the above method, the step (a) may include a step of selecting one color chart data to be used for creating the color prediction model from a plurality of color chart data created in advance. This method allows the selection of appropriate color chart data for use in creating a color prediction model.

[0082] (7) In the above method, step (a) may include a step of displaying a list indicating the plurality of color chart data, and the list may include, as setting information suitable for each color chart data, printing medium information indicating one printing medium among a plurality of printing media usable by the printing device, and ink set information indicating one ink set among a plurality of ink sets usable by the printing device. According to this method, it is possible to select color chart data suitable for the combination of printing medium and ink set as the color chart data used to create the color prediction model.

[0083] (8) According to a second aspect of the present disclosure, there is provided a color prediction model creation device that creates a color prediction model that predicts spectral reflectance from an ink amount set of multiple types of inks that constitute an ink set. The color prediction model creation device includes a data acquisition unit that acquires color chart data representing N color patch groups, where N is an integer equal to or greater than 2, an accuracy calculation unit that sequentially executes an accuracy calculation process while increasing the integer n, the accuracy calculation unit performing the following steps: (i) creating n sets of temporary learning data sets corresponding to the n color patch groups using predicted spectral reflectances predicted by a simplified color prediction model that can be learned more easily than the color prediction model, (ii) creating a temporary color prediction model by learning the color prediction model using the n sets of temporary learning data sets, and (iii) calculating the prediction accuracy of the temporary color prediction model; and (ii) calculating the number n of the color patch groups suitable for learning the color prediction model from the relationship between the integer n and the prediction accuracy. c A determination unit that determines c The system includes a learning data creation unit that performs printing using a printing device and measuring the spectral reflectance of the color patch groups, and creates learning data including the measurement results of the spectral reflectance, and a learning unit that uses the learning data to learn the color prediction model.

[0084] (9) According to a third aspect of the present disclosure, a printing system is provided. The printing system includes a color prediction model creation device that creates a color prediction model that predicts spectral reflectance from an ink amount set of multiple types of ink that constitute an ink set, and a printing device. The color prediction model creation device includes a data acquisition unit that acquires color chart data representing N color patch groups, where N is an integer equal to or greater than 2, an accuracy calculation unit that sequentially executes an accuracy calculation process while increasing the integer n, the accuracy calculation unit performing the following steps, where n is an integer equal to or greater than 1 and equal to or less than N: (i) create n sets of temporary learning data sets corresponding to the n color patch groups using predicted spectral reflectances predicted by a simplified color prediction model that can be learned more easily than the color prediction model, (ii) create a temporary color prediction model by learning the color prediction model using the n sets of temporary learning data sets, and (iii) calculate the prediction accuracy of the temporary color prediction model; and (ii) calculate the number n of the color patch groups suitable for learning the color prediction model from the relationship between the integer n and the prediction accuracy. c A determination unit that determines c The color prediction model includes a learning data creation unit that performs printing by the printing device and measurement of spectral reflectance for each of the color patch groups, and creates learning data including the measurement results of the spectral reflectance, and a learning unit that uses the learning data to learn the color prediction model.

[0085] (10) According to a fourth aspect of the present disclosure, there is provided a computer program for creating a color prediction model that predicts spectral reflectance from an ink amount set of a plurality of types of ink that constitute an ink set. This computer program includes: (a) a process for acquiring color chart data representing N color patch groups, where N is an integer equal to or greater than 2; (b) a process for sequentially executing, while increasing the integer n, a calculation process for calculating the number n of color patch groups suitable for training the color prediction model, where n is an integer equal to or greater than 1 and equal to or less than N, the following steps: (i) creating n sets of provisional training data sets corresponding to the n color patch groups using predicted spectral reflectances predicted by a simplified color prediction model that can be trained more easily than the color prediction model, (ii) training the color prediction model using the n sets of provisional training data sets, thereby creating a provisional color prediction model; and (iii) calculating the prediction accuracy of the provisional color prediction model, where n is an integer equal to or greater than 1 and equal to or less than N, the calculation process for calculating the number n of color patch groups suitable for training the color prediction model, while increasing the integer n; and (c) calculating the number n of color patch groups suitable for training the color prediction model from the relationship between the integer n and the prediction accuracy. c (d) determining said n c The computer is caused to execute a process of (a) printing the color patch groups using a printing device, measuring the spectral reflectance, and creating learning data including the results of the measurement of the spectral reflectance, and (b) learning the color prediction model using the learning data.

[0086] The present disclosure may be realized in various forms other than an image processing device, a printing system, and a computer program, for example, an image processing method, a non-transitory storage medium on which a computer program is recorded, and the like. [Explanation of symbols]

[0087] 50...CPU, 51...processing setting unit, 52...data acquisition unit, 53...accuracy calculation unit, 54...determination unit, 55...learning data creation unit, 56...learning unit, 60...storage unit, 61...color prediction model creation program, 62...color chart data, 63...color chart information, 64...accuracy confirmation data, 70...input / output interface, 100...color prediction model creation device, 200...printing device, 210...ink container, 220...head scanning mechanism, 300...color measurement device, 400...display device, 500...printing system

Claims

1. A method for creating a color prediction model that predicts a spectral reflectance from an ink amount set of a plurality of types of ink that constitute an ink set, comprising the steps of: (a) acquiring color chart data representing a group of N color patches, where N is an integer equal to or greater than 2; (b) sequentially executing an accuracy calculation process, where n is an integer between 1 and N, inclusive, while increasing the integer n, in which: (i) n sets of temporary learning data sets corresponding to n sets of color patches are created using predicted spectral reflectances predicted by a simplified color prediction model that can be learned more easily than the color prediction model; (ii) a temporary color prediction model is created by learning the color prediction model using the n sets of temporary learning data sets; and (iii) a prediction accuracy of the temporary color prediction model is calculated. (c) determining the number n of the color patch sets suitable for learning the color prediction model from the relationship between the integer n and the prediction accuracy; ｃ determining (d) the n ｃ a step of printing the color patch group by a printing device and measuring the spectral reflectance thereof, and creating learning data including the measurement results of the spectral reflectance; (e) performing training of the color prediction model using the training data; A method comprising:

2. 2. The method of claim 1 , The step (b) includes a step of stopping the increment of the integer n when the prediction accuracy reaches a target accuracy or less; The step (c) is to set the integer n when the prediction accuracy reaches or is equal to or lower than the target accuracy, to the number n of the color patch groups. ｃ The method of claim 1, further comprising the step of determining

3. 2. The method of claim 1 , The step (b) is continuously performed until the integer n reaches the integer N; The step (c) (c1) displaying a relationship between the integer n and the prediction accuracy; (c2) The number n of the color patch groups by the user ｃ accepting a designation of A method comprising:

4. 2. The method of claim 1 , The step (a) includes a step of sequentially selecting one color chart data to be used for creating the color prediction model from a plurality of color chart data created in advance, The step (b) is performed for each of the selected color chart data, The step (c) is a step of: determining the number n of the color patch groups for one color chart data having the best prediction accuracy among the plurality of color chart data; ｃ The method includes determining

5. 2. The method of claim 1 , A method according to claim 1, wherein one of said color patch groups is a color chart printed on a sheet of printing medium.

6. 2. The method of claim 1 , The method, wherein the step (a) includes a step of selecting one color chart data set to be used in creating the color prediction model from a plurality of color chart data sets created in advance.

7. 7. The method of claim 6, The step (a) includes a step of displaying a list showing the plurality of color chart data, The method, wherein the list includes, as setting information suitable for each color chart data, print medium information indicating one print medium among a plurality of print media usable by the printing device, and ink set information indicating one ink set among a plurality of ink sets usable by the printing device.

8. A color prediction model creation device that creates a color prediction model that predicts a spectral reflectance from an ink amount set of a plurality of types of ink that constitute an ink set, comprising: a data acquisition unit that acquires color chart data representing a group of N color patches, where N is an integer equal to or greater than 2; an accuracy calculation unit that sequentially executes an accuracy calculation process while increasing the integer n, the accuracy calculation process being as follows: (i) creating n sets of temporary learning data sets corresponding to n sets of color patches using predicted spectral reflectances predicted by a simplified color prediction model that can be learned more easily than the color prediction model, (ii) creating a temporary color prediction model by learning the color prediction model using the n sets of temporary learning data sets, and (iii) calculating a prediction accuracy of the temporary color prediction model, where n is an integer between 1 and N inclusive; From the relationship between the integer n and the prediction accuracy, the number n of the color patch groups suitable for learning the color prediction model is determined. ｃ A determination unit for determining The n ｃ a learning data creation unit that executes printing by a printing device and measurement of spectral reflectance for the color patch group, and creates learning data including the measurement results of the spectral reflectance; a learning unit that uses the learning data to learn the color prediction model; A color prediction model creating device comprising:

9. 1. A printing system comprising: a color prediction model creating device that creates a color prediction model that predicts a spectral reflectance from an ink amount set of a plurality of types of ink that constitute an ink set; A printing device; Equipped with The color prediction model creating device comprises: a data acquisition unit that acquires color chart data representing a group of N color patches, where N is an integer equal to or greater than 2; an accuracy calculation unit that sequentially executes an accuracy calculation process while increasing the integer n, the accuracy calculation process being as follows: (i) creating n sets of temporary learning data sets corresponding to n sets of color patches using predicted spectral reflectances predicted by a simplified color prediction model that can be learned more easily than the color prediction model, (ii) creating a temporary color prediction model by learning the color prediction model using the n sets of temporary learning data sets, and (iii) calculating a prediction accuracy of the temporary color prediction model, where n is an integer between 1 and N inclusive; From the relationship between the integer n and the prediction accuracy, the number n of the color patch groups suitable for learning the color prediction model is determined. ｃ A determination unit for determining The n ｃ a learning data creation unit that executes printing by the printing device and measurement of spectral reflectance for each of the color patch groups, and creates learning data including the measurement results of the spectral reflectance; a learning unit that uses the learning data to learn the color prediction model; A printing system comprising:

10. A computer program for creating a color prediction model for predicting a spectral reflectance from an ink amount set of a plurality of types of ink constituting an ink set, comprising: (a) acquiring color chart data representing a group of N color patches, where N is an integer equal to or greater than 2; (b) a process of sequentially executing an accuracy calculation process, in which n is an integer between 1 and N, while increasing the integer n, in which: (i) n sets of temporary learning data sets corresponding to n sets of color patches are created using predicted spectral reflectances predicted by a simplified color prediction model that can be learned more easily than the color prediction model; (ii) a temporary color prediction model is created by learning the color prediction model using the n sets of temporary learning data sets; and (iii) a prediction accuracy of the temporary color prediction model is calculated. (c) determining the number n of the color patch sets suitable for learning the color prediction model from the relationship between the integer n and the prediction accuracy; ｃ and determining (d) the n ｃ a process of performing printing by a printing device and measuring the spectral reflectance of each of the color patch groups, and creating learning data including the measurement results of the spectral reflectance; (e) performing learning of the color prediction model using the learning data; and A computer program that causes a computer to execute the following.

Images