Ai-based color mixing device and operating method thereof

MYPI2025004951A0Undetermined Publication Date: 2024-09-02KCC CORP
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
MYPI2025004951
Authority / Receiving Office
MY · MY
Patent Type
Applications
Current Assignee / Owner
Priority Date
2023-03-02
Filing Date
2024-01-04
Publication Date
2024-09-02

AI Technical Summary

Technical Problem

Current color mixing technologies, such as CCM, often result in significant color differences between the combined color and the target color, requiring manual correction by operators, which increases processing time and depends on their skill level, leading to inefficiencies and potential failure in achieving the target color.

Method used

An AI-based color mixing device and method that determines the optimal mixing conditions by selecting an appropriate artificial intelligence model based on the target color's location in the color space, generating samples within defined mixing ratios, and minimizing color difference values, using a deep learning model to predict the most accurate mixing ratios for reference colors.

Benefits of technology

This approach significantly reduces the time required for color mixing and improves accuracy by optimizing the mixing process, ensuring the target color is achieved with reduced operator intervention and skill dependency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure PI2025004951_DRAWING_001
    Figure PI2025004951_DRAWING_001
  • Figure PI2025004951_DRAWING_002
    Figure PI2025004951_DRAWING_002
  • Figure PI2025004951_DRAWING_003
    Figure PI2025004951_DRAWING_003
Patent Text Reader

Abstract

According to an embodiment disclosed in the present disclosure, an operating method of a color mixing device includes determining a region, which includes a color value of a target color, from among one or more regions included in a color space based on the color value of the target color, generating one or more samples based on the color value of the target color, selecting a deep learning model for the determined region from a plurality of pre-learned AI models, and determining a sample, which is an optimal mixing condition for composing the color value of the target color, from among the samples by using the selected AI model. The plurality of AI models are learned based on learning-specific data sets in which a mixing ratio of each of one or more reference colors is arbitrarily set.
Need to check novelty before this filing date? Find Prior Art

Description

AI-based color mixing device and its operation method

[0001] Cross-citation with related applications

[0002] This invention claims the benefit of priority from Korean Patent Application No. 10-2023-0028019, filed on March 2, 2023, and all contents of the document in that Korean Patent Application are incorporated herein by reference.

[0003] Technology field

[0004] Embodiments disclosed in this document relate to an AI-based color mixing device and an operating method thereof.

[0005] Computer Color Matching (CCM) is a technology used to determine paint color during the color mixing process. CCM measures the reflectance of a target color and can derive mixing conditions, including the mixing ratios of each reference color (e.g., red, green, and blue). If the resulting color differs from the target color, the color must be corrected by the operator. Therefore, accurate mixing conditions are essential.

[0006] When determining color combinations using the Color Matching Model (CCM), significant color differences can occur between the colors combined through the CCM and the target color. To reduce this difference, the artist must perform additional work to correct the combined colors.

[0007] Accordingly, there is a problem that the time required to perform color mixing increases due to additional work after color mixing using CCM.

[0008] Additionally, there is a problem that whether the target color can be achieved or not may depend on the experience and skill of the worker.

[0009] The technical problems of the embodiments disclosed in this document are not limited to the technical problems mentioned above, and other technical problems not mentioned will be clearly understood by those skilled in the art from the descriptions below.

[0010] An operating method of a color mixing device according to an embodiment disclosed in the present document may include: an operation of determining, based on a color value of a target color, an area including a color value of the target color among one or more areas included in a color space; an operation of generating one or more samples based on the color value of the target color; an operation of selecting an artificial intelligence model for the determined area among a plurality of previously learned artificial intelligence models; and an operation of determining, using the selected artificial intelligence model, a sample having an optimal mixing condition for composing a color value of the target color among the samples, wherein the plurality of artificial intelligence models may be learned based on learning data sets in which a mixing ratio of each of one or more reference colors is arbitrarily set.

[0011] In one embodiment, the operation of generating the samples may generate the samples based on a primary mixing condition in which the reference colors are randomly mixed within a minimum mixing ratio or a maximum mixing ratio for each of the reference colors constituting the determined area.

[0012] In one embodiment, the operation of generating the samples may generate the samples based on secondary mixing conditions that arbitrarily change the primary mixing conditions within a preset range.

[0013] In one embodiment, the determining operation may determine a sample having a color difference value between the color values ​​of the samples and the color value of the target color less than a preset value as an optimal mixing condition.

[0014] In one embodiment, the determining operation may determine the optimal mixing condition as a sample in which the color difference value between the color values ​​of the samples and the color value of the target color is minimized.

[0015] According to an embodiment disclosed in the present document, a color mixing device may include: a region determining unit that determines a region including a color value of a target color among one or more regions included in a color space based on a color value of the target color; a sample generating unit that generates one or more samples based on the color value of the target color; a model selecting unit that selects an artificial intelligence model for the determined region among a plurality of previously learned artificial intelligence models; and a determining unit that determines a sample having an optimal mixing condition for composing a color value of the target color among the samples using the selected artificial intelligence model, wherein the plurality of artificial intelligence models may be learned based on learning data sets in which a mixing ratio of each of one or more reference colors is arbitrarily set.

[0016] In one embodiment, the sample generation unit can generate the samples based on a primary mixing condition in which the reference colors are randomly mixed within a minimum mixing ratio or a maximum mixing ratio for each of the reference colors constituting the determined area.

[0017] In one embodiment, the sample generation unit can generate the samples based on secondary mixing conditions that are arbitrarily changed within a preset range of the primary mixing conditions.

[0018] In one embodiment, the decision unit can determine a sample having a color difference value between the color values ​​of the samples and the color value of the target color less than a preset value as an optimal mixing condition.

[0019] In one embodiment, the decision unit can determine the optimal mixing condition as a sample in which the color difference value between the color values ​​of the samples and the color value of the target color is minimized.

[0020] The color combination determination device and its operating method according to various embodiments disclosed in this document can divide a color space into a plurality of regions and build an artificial intelligence model optimized for a specific color region.

[0021] The color combination determination device and its operating method according to various embodiments disclosed in this document can select an optimized artificial intelligence model among a plurality of artificial intelligence models based on the color value of a target color.

[0022] Accordingly, the time for the color mixing process can be shortened and the accuracy of mixing can be improved.

[0023] The effects of the color mixing determination device and the operating method thereof according to the disclosure of this document are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art according to the disclosure of this document.

[0024] FIG. 1 is a block diagram of a color mixing device according to an embodiment of the present disclosure.

[0025] FIG. 2 illustrates a color space based on the Lab color system according to one embodiment of the present disclosure.

[0026] FIG. 3 illustrates a table of reference colors by region according to one embodiment of the present disclosure.

[0027] FIG. 4 is a flowchart illustrating a deep learning model training method according to an embodiment of the present disclosure.

[0028] FIG. 5 is a flowchart illustrating an operation method of a color mixing device according to an embodiment of the present disclosure.

[0029] In connection with the description of the drawings, the same or similar reference numerals may be used for identical or similar components.

[0030] Hereinafter, embodiments of the present invention will be described with reference to the attached drawings. However, this is not intended to limit the present invention to specific embodiments, and it should be understood that the present invention encompasses various modifications, equivalents, and / or alternatives of the embodiments.

[0031] The embodiments and terminology used in this document are not intended to limit the technical features described in this document to a specific embodiment, but should be understood to encompass various modifications, equivalents, or alternatives of the embodiment. In connection with the description of the drawings, similar reference numerals may be used to refer to similar or related components. The singular form of a noun corresponding to an item may include one or more of the item, unless the relevant context clearly indicates otherwise.

[0032] In this document, the phrases "A or B", "at least one of A and B", "at least one of A or B", "A, B, or C", "at least one of A, B, and C", and "at least one of A, B, or C" can each include any one of the items listed together in that phrase, or all possible combinations thereof. Terms such as "first", "second", "first", "second", "A", "B", "(a)", or "(b)" may be used merely to distinguish the corresponding component from other corresponding components, and do not limit the corresponding components in any other respect (e.g., importance or order) unless specifically stated otherwise.

[0033] In this document, when a component (e.g., a first component) is referred to as being “connected,” “coupled,” or “connected,” with or without the terms “functionally” or “communicatively,” or “coupled” or “connected,” it means that the component can be connected to the other component directly (e.g., wired or wirelessly), or indirectly (e.g., via a third component).

[0034] The methods according to various embodiments disclosed in this document may be provided as included in a computer program product. The computer program product may be traded as a commodity between a seller and a buyer. The computer program product may be distributed in the form of a machine-readable storage medium (e.g., compact disc read-only memory, CD-ROM), or may be distributed online (e.g., downloaded or uploaded) through an application store or directly between two user devices. In the case of online distribution, at least a portion of the computer program product may be temporarily stored or temporarily generated in a machine-readable storage medium, such as the memory of a manufacturer's server, an application store's server, or an intermediary server.

[0035] According to the embodiments disclosed in this document, each component (e.g., a module or a program) of the above-described components may include one or more entities, and some of the entities may be separated and placed in other components. According to the embodiments disclosed in this document, one or more components or operations of the aforementioned components may be omitted, or one or more other components or operations may be added. Alternatively or additionally, a plurality of components (e.g., a module or a program) may be integrated into a single component. In this case, the integrated component may perform one or more functions of each of the plurality of components identically or similarly to those performed by the corresponding component among the plurality of components prior to the integration. According to the embodiments disclosed in this document, the operations performed by a module, program, or other component may be executed sequentially, in parallel, iteratively, or heuristically, or one or more of the operations may be executed in a different order, omitted, or one or more other operations may be added.

[0036] FIG. 1 is a block diagram of a color mixing device (10) according to one embodiment of the present disclosure.

[0037] Referring to FIG. 1, the color mixing device (10) may include a sensor (120), a memory (140), and a processor (160), or a combination thereof. According to an embodiment, the color mixing device (10) illustrated in FIG. 1 may further include at least one component (e.g., a display, an input device, or an output device) other than the components illustrated in FIG. 1.

[0038] In one embodiment, the sensor (120) may be an optical sensor. In one embodiment, the sensor (120) may detect the reflectance of a color. For example, the sensor (120) may express the color value of the detected color as a Lab color value based on the reflectance of the detected color.

[0039] In one embodiment, the memory (140) may include volatile memory and / or non-volatile memory.

[0040] In one embodiment, the memory (140) may store data used by at least one component (e.g., processor (160)) of the color mixing device (10). For example, the data may include software (or instructions related thereto), input data, or output data. In one embodiment, the instructions, when executed by the processor (160), may cause the color mixing device (10) to perform operations defined by the instructions.

[0041] In one embodiment, the memory (140) may include one or more software (e.g., a region determination unit (141), a sample generation unit (143), a model learning unit (145), a model selection unit (147), a determination unit (149), or a combination thereof).

[0042] In one embodiment, the processor (160) may include a central processing unit, an application processor, a graphics processing unit, a neural processing unit (NPU), an image signal processor, a sensor hub processor, or a communication processor.

[0043] In one embodiment, the processor (160) may execute software (e.g., a region determination unit (141), a sample generation unit (143), a model learning unit (145), a model selection unit (147), a determination unit (149), or a combination thereof) to control at least one other component (e.g., a hardware or software component) of a color mixing device (10) connected to the processor (160) and perform various data processing or calculations.

[0044] In one embodiment, the database (15) may include data regarding color values ​​according to color codes, color mixing data for previously performed color mixing processes, and data regarding minimum and maximum mixing ratios for each of the reference colors.

[0045] Hereinafter, a method for a color mixing device (10) to determine optimal mixing conditions for a target color through a region determination unit (141), a sample generation unit (143), a model learning unit (145), a model selection unit (147), a determination unit (149), or a combination thereof is described.

[0046] AI model training method

[0047] The model learning unit (145) can receive learning data sets. The model learning unit (145) can receive learning data sets from the database (15). Here, the learning data set may be data acquired during a previously performed coloring process. The learning data set may include data regarding a color code of a specific color, a color value of a specific color, mixing conditions for mixing a specific color, or a combination thereof.

[0048] The model learning unit (145) can distinguish training data sets based on specific criteria. The model learning unit (145) can distinguish training data sets by region in a color space. Here, the color space is a space that expresses a color system in three dimensions, and may be a set of colors based on the RGB (red, green, blue) color system or the Lab color system. Colors included in the Lab color system can be expressed through a combination of lightness, red-green series (a channel), and yellow-blue series (b channel). The reference colors can be expressed by a combination of reference colors. The reference colors can be primary colors. The reference colors of the Lab color system can include black and white representing lightness, red and green representing a channel, and yellow and blue representing b channel. A plurality of regions can be configured based on reference colors (e.g., black or white, red or green, yellow or blue). For example, if a color space is divided into eight regions, the first region among the eight regions can be configured based on white, red, and yellow, and the colors included in the first region can be expressed using white, red, and yellow as reference colors.

[0049] In the following, it is assumed that the colors are Lab color systems and that the number of regions in the color space is eight. However, this is merely a premise for explanation, and the present invention is not limited to the Lab color system, but can also be applied to various other color systems (e.g., RGB color system), and the number of regions in the color space can also be divided into various numbers of regions.

[0050] The model learning unit (145) can train one or more artificial intelligence models. The model learning unit (145) can train an artificial intelligence model for each domain. The model learning unit (145) can train one or more artificial intelligence models using training data sets. The model learning unit (145) can train an artificial intelligence model for each domain using training data sets classified by domain. Here, the artificial intelligence model for each domain may be an artificial intelligence model optimized for each domain. The artificial intelligence models may include machine learning models and deep learning models.

[0051] In one embodiment, the model learning unit (145) may train one or more artificial intelligence models using training data sets based on supervised learning. For example, the model learning unit (145) may train the artificial intelligence model to output mixing conditions for mixing a specific color when the color value of a specific color included in the training data set is input to the artificial intelligence model. Here, the mixing conditions may be conditions that include all information on the mixing ratio and mixing amount of each of the reference colors in order to mix a specific color using the reference colors.

[0052] Artificial Intelligence Model R 2(Coefficient of determination)Prediction accuracy (color difference value)Linear Regression0.78214.05Lasso Regression0.75514.86K-Nearest Neighbor0.8699.83Decision Tree0.9316.98Support Vector Regression0.83611.43Ada Boost0.83212.86Gradient Boost Regression0.9387.16XG Boost Regression0.9803.79Random Forest Regression0.9694.45Extra Tree Regression0.9793.55Neural Network Regression0.9971.458 combination NN Regression0.9990.78

[0053] Table 1 illustrates various AI models. Here, the coefficient of determination can be an indicator of model fit, and prediction accuracy can be an indicator of the degree of similarity between the color value determined using the AI ​​model and the color value of the target color.

[0054] Referring to Table 1, the artificial intelligence model trained by the model learning unit (145) may be a deep learning model implementing NN (neural network) Regression. The combination of deep learning models for each of the eight areas may be referred to as 8-combination NN Regression.

[0055] Referring to Table 1, the present invention can determine the mixing sample closest to the color value of the target color using 8-combination NN Regression. When using 8-combination NN Regression, it can be confirmed that compared to other artificial intelligence models, the coefficient of determination is high, the model fit is high, and the color difference value is low, allowing for the most accurate prediction of the color value of the target color.

[0056] Below, the AI ​​model is assumed to be a deep learning model with three layers, each with a node count of (64-64-40). However, this is merely a premise for illustrative purposes, and the present invention is not limited to deep learning models and can be applied to various other AI models.

[0057] How to determine optimal mixing conditions

[0058] The region determination unit (141) can determine one region among one or more regions based on the color value of the target color. The region determination unit (141) can determine a region including the color value of the target color among one or more regions included in the Lab color space based on the color value of the target color. For example, when the color value of the target color is expressed as white, red, and yellow, the region determination unit (141) can determine a first region among a plurality of regions with white, red, and yellow as the reference color as the region including the target color.

[0059] The region determination unit (141) can search for the color value of the target color in order to determine a region including the color value of the target color. The region determination unit (141) can search for the color value of the target color based on data stored in the database (15). The region determination unit (141) can search for the color value of the target color based on the color code of the target color. The region determination unit (141) can search for the color value of the target color by comparing the color code of the target color with the data stored in the database (15). Here, the color value may be a value for specifying a specific color. For example, the color value in the Lab color system may be a coordinate on the Lab color space. The color value in the Lab color system may be a coordinate based on the x-axis (a channel), the y-axis (b channel), and the z-axis (brightness) on the Lab color space. In one embodiment, the color code may mean information indicating the target color in a classification standard for color distinction different from Lab.

[0060] The sample generation unit (143) can generate one or more samples based on the color value of the target color. Here, the sample may be a color combined based on one of a plurality of mixing conditions for configuring the color value of the target color.

[0061] In one embodiment, the sample generation unit (143) may generate one or more samples using a screening technique. The sample generation unit (143) may generate one or more samples using a primary screening technique and / or a secondary screening technique. Here, the primary screening technique may be a technique for generating samples based on macroscopic mixing conditions for mixing a target color. The secondary screening technique may be a technique for generating samples by arbitrarily changing the macroscopic mixing conditions within a preset range.

[0062] The sample generation unit (143) can generate one or more samples using a primary screening technique. The sample generation unit (143) can generate one or more samples based on macroscopic mixing conditions. The sample generation unit (143) can generate one or more samples based on the primary mixing conditions. The sample generation unit (143) can generate samples by randomly combining one or more reference colors within a minimum mixing ratio to a maximum mixing ratio of each of the reference colors. The sample generation unit (143) can generate one or more samples based on the primary mixing conditions in which the reference colors are randomly mixed within a minimum mixing ratio to a maximum mixing ratio of each of the reference colors. Here, the minimum mixing ratio and the maximum mixing ratio can be set based on statistical data of mixing ratios for the reference colors in a previously performed color mixing process. For example, if the average mixing ratio of white, which is the reference color, is 30% in the previously performed coloring process, the minimum mixing ratio of white can be set to 10% and the maximum mixing ratio can be set to 50%.

[0063] The sample generation unit (143) can generate samples using a secondary screening technique. The sample generation unit (143) can derive macroscopic mixing conditions using the primary screening technique, and can generate samples according to mixing conditions within a preset range from the macroscopic mixing conditions using the secondary screening technique. The sample generation unit (143) can generate one or more samples according to secondary mixing conditions in which the primary mixing conditions are arbitrarily changed within a preset range. Here, the preset range can be set by a setter. For example, when the maximum mixing ratio of white, which is a reference color, is 50% and the mixing ratio of white included in the macroscopic mixing conditions is 50%, the sample generation unit (143) can generate samples in which the maximum mixing ratio of white is changed within 10% to 30%.

[0064] The sample generation unit (143) can quickly generate samples close to the color value of the target color based on macroscopic mixing conditions by using the primary screening and / or secondary screening techniques.

[0065] The model selection unit (147) can select at least one deep learning model from among multiple pre-trained deep learning models. The model selection unit (147) can select at least one deep learning model from among multiple pre-trained deep learning models based on the color value of the target color.

[0066] The decision unit (149) can determine optimal mixing conditions for configuring color values ​​of a target color using the deep learning model selected by the model selection unit (147). In one embodiment, the decision unit (149) can determine a sample corresponding to the optimal mixing conditions among a plurality of samples using the deep learning model.

[0067] The decision unit (149) can use the color difference value between the color values ​​of the samples and the color value of the target color to determine the optimal mixing conditions among the samples. The color difference value can be expressed by mathematical expression 1.

[0068]

[0069] Here, ΔE can mean a chrominance value. L, a, and b each mean a coordinate component in the Lab color space, the subscript 1 can mean a coordinate component for the color value of the sample, and the subscript 2 can mean a coordinate component for the color value of the target color.

[0070] In one embodiment, the decision unit (149) can determine a sample among a plurality of samples whose color difference value is less than a preset value as the optimal mixing condition.

[0071] In one embodiment, the decision unit (149) can determine the sample with the smallest color difference value among the plurality of samples as the optimal mixing condition.

[0072] FIG. 2 illustrates a color space (200) based on the Lab color system according to one embodiment of the present disclosure.

[0073] Referring to FIG. 2, the color space (200) may be a spherical space. The color space (200) may include reference lines (210, 212, 214). Here, the reference lines (210, 212, 214) may be lines connecting the coordinates of reference colors. The color space (200) may be divided into eight regions by the reference lines (210, 212, 214). The color space (200) may be divided into eight regions with the reference lines (210, 212, 214) as boundaries. For example, a portion of a color space (200) defined by a line segment of a first reference line (210) toward white (L=100), a line segment of a second reference line (212) toward red (+a), and a line segment of a third reference line (214) toward yellow (+b) with respect to the origin (202) as the boundary may be referred to as a first region. In addition, a portion of a color space (200) defined by a line segment of a first reference line (210) toward white (L=100), a line segment of a second reference line (212) toward red (+a), and a line segment of a third reference line (214) toward blue (-b) with respect to the origin (202) as the boundary may be referred to as a second region. The reference color of the first region may be white, red, yellow, or a combination thereof, and the reference color of the second region may be white, red, blue, or a combination thereof.

[0074] FIG. 3 illustrates a table (300) of reference colors by region according to one embodiment of the present disclosure.

[0075] Referring to FIG. 3, the color space (200) may be divided into one or more regions. For example, the color space (200) may be divided into eight regions by reference lines (210, 212, 214). Each of the eight regions has a reference color for expressing the colors included in each region.

[0076] Referring to table (300), the reference color of the first region may be white, red, yellow, or a combination thereof. The colors included in the first region may be expressed through color values ​​of white, red, yellow, or a combination thereof. Here, the color values ​​may be coordinates on the color space (200). For example, the color value of white may be (0, 0, 100), the color value of red may be (a, 0, 0), and the color value of yellow may be (0, b, 0). When the target color is included in the first region, the color value of the target color may be expressed based on the color values ​​of white, red, and yellow.

[0077] FIG. 4 is a flowchart illustrating a deep learning model training method according to an embodiment of the present disclosure.

[0078] Referring to FIG. 4, in operation 400, the model learning unit (145) can distinguish training data sets based on specific criteria. The model learning unit (145) can distinguish training data sets by region in the color space.

[0079] In operation 402, the model learning unit (145) can train one or more deep learning models. The model learning unit (145) can train a deep learning model for each domain. The model learning unit (145) can train one or more deep learning models using training data sets. The model learning unit (145) can train a deep learning model for each domain using training data sets classified by domain.

[0080] In one embodiment, the model learning unit (145) can train one or more deep learning models using learning data sets based on supervised learning.

[0081] FIG. 5 is a flowchart showing an operation method of a color mixing device (10) according to one embodiment of the present disclosure.

[0082] Referring to FIG. 5, in operation 500, the region determination unit (141) may determine one of one or more regions based on the color value of the target color. The region determination unit (141) may determine a region including the color value of the target color among one or more regions included in the Lab color space based on the color value of the target color.

[0083] The region determination unit (141) can search for the color value of the target color in order to determine the region including the color value of the target color. The region determination unit (141) can search for the color value of the target color based on data stored in the database (15). The region determination unit (141) can search for the color value of the target color based on the color code of the target color. The region determination unit (141) can search for the color value of the target color by comparing the color code of the target color with data stored in the database (15).

[0084] In operation 502, the sample generation unit (143) can generate one or more samples based on the color value of the target color.

[0085] In one embodiment, the sample generation unit (143) may generate one or more samples using a screening technique. The sample generation unit (143) may generate one or more samples using a primary screening technique and / or a secondary screening technique.

[0086] The sample generation unit (143) can generate one or more samples using a primary screening technique. The sample generation unit (143) can generate one or more samples based on macroscopic mixing conditions. The sample generation unit (143) can generate samples by arbitrarily combining one or more reference colors within a minimum mixing ratio or a maximum mixing ratio of each reference color.

[0087] The sample generation unit (143) can generate samples using a secondary screening technique. The sample generation unit (143) can derive macroscopic mixing conditions using a primary screening technique, and generate samples according to mixing conditions within a preset range from the macroscopic mixing conditions using a secondary screening technique.

[0088] In operation 504, the model selection unit (147) may select at least one deep learning model from among a plurality of pre-trained deep learning models. The model selection unit (147) may select at least one deep learning model from among a plurality of pre-trained deep learning models based on the color value of the target color.

[0089] In operation 506, the decision unit (149) may determine optimal mixing conditions for configuring color values ​​of a target color using the deep learning model selected by the model selection unit (147). In one embodiment, the decision unit (149) may determine a sample corresponding to the optimal mixing conditions among a plurality of samples using the deep learning model.

[0090] The decision unit (149) can use the color difference value of the color values ​​of the samples and the target color value to determine the optimal mixing conditions among the samples.

[0091] In one embodiment, the decision unit (149) can determine a sample among a plurality of samples whose color difference value is less than a preset value as the optimal mixing condition.

[0092] In one embodiment, the decision unit (149) can determine the sample with the smallest color difference value among the plurality of samples as the optimal mixing condition.

[0093] The terms "include," "comprise," or "have" used herein, unless otherwise specifically stated, imply that the corresponding component may be included, and therefore should be interpreted to include other components rather than to exclude other components. All terms, including technical or scientific terms, have the same meaning as commonly understood by a person of ordinary skill in the art to which the embodiments disclosed in this document pertain, unless otherwise defined. Commonly used terms, such as terms defined in dictionaries, should be interpreted to be consistent with their contextual meaning in the relevant art, and shall not be interpreted in an idealized or overly formal sense, unless explicitly defined in this document.

[0094] The above description is merely an example of the technical idea disclosed in this document, and those skilled in the art to which the embodiments disclosed in this document pertain may make various modifications and variations without departing from the essential characteristics of the embodiments disclosed in this document. Therefore, the embodiments disclosed in this document are not intended to limit the technical idea of ​​the embodiments disclosed in this document, but to explain it, and the scope of the technical idea disclosed in this document is not limited by these embodiments. The scope of protection of the technical idea disclosed in this document should be interpreted by the claims below, and all technical ideas within a scope equivalent thereto should be interpreted as being included in the scope of rights of this document.

Claims

1. An operation of determining an area including a color value of the target color among one or more areas included in a color space based on a color value of the target color; An operation of generating one or more samples based on the color value of the target color; An operation of selecting a deep learning model for the determined area among multiple learned artificial intelligence models; and An operation of determining a sample that is an optimal mixing condition for composing a color value of the target color among the samples using the above-mentioned selected artificial intelligence model is included. The above multiple artificial intelligence models are trained based on training data sets in which the mixing ratio of each of one or more reference colors is arbitrarily set. How the color mixing device works.

2. In claim 1, The action of generating the above samples is: Generating the samples based on the primary mixing conditions in which the reference colors are randomly mixed within the minimum mixing ratio and maximum mixing ratio for each of the reference colors constituting the determined area. How it works.

3. In claim 2, The action of generating the above samples is: The samples are generated based on the secondary mixing conditions in which the primary mixing conditions are arbitrarily changed within a preset range. How it works.

4. In claim 1, The above decision-making action is, The optimal mixing conditions are determined by determining a sample in which the color difference value between the color values ​​of the above samples and the color value of the target color is less than a preset value. How it works.

5. In claim 1, The above decision-making action is, The optimal mixing conditions are determined by minimizing the color difference value between the color values ​​of the above samples and the color value of the target color. How it works.

6. A region determination unit that determines a region among one or more regions included in a color space based on a color value of the target color, the region including the color value of the target color; A sample generation unit for generating one or more samples based on the color value of the target color; A model selection unit that selects an artificial intelligence model for the determined area among multiple learned artificial intelligence models; and Using the above-mentioned selected artificial intelligence model, a decision unit is included that determines a sample that is an optimal mixing condition for composing a color value of the target color among the samples, The above multiple artificial intelligence models are trained based on training data sets in which the mixing ratio of each of one or more reference colors is arbitrarily set. Color mixing device.

7. In claim 6, The above sample generation unit, Generating the samples based on the primary mixing conditions in which the reference colors are randomly mixed within the minimum mixing ratio and maximum mixing ratio for each of the reference colors constituting the determined area. Color mixing device.

8. In claim 7, The above sample generation unit, The samples are generated based on the secondary mixing conditions in which the primary mixing conditions are arbitrarily changed within a preset range. Color mixing device.

9. In claim 6, The above decision-making body, The optimal mixing conditions are determined by determining a sample in which the color difference value between the color values ​​of the above samples and the color value of the target color is less than a preset value. Color mixing device.

10. In claim 6, The above decision-making body, The optimal mixing conditions are determined by minimizing the color difference value between the color values ​​of the above samples and the color value of the target color. Color mixing device.