High tone color assistance program, system, method, and machine learning model

Abstract

Provided is a bleaching condition necessary for making the hair after hair dyeing a high tone color.　Specifically provided is a high tone color assistance program that is characterized by causing a computer to function as a hair damage level acquisition means, a pre-bleaching tone acquisition means, a target post-dyeing tone acquisition means, and a bleaching condition acquisition means. The high tone color assistance program is also characterized in that the bleaching condition acquisition means inputs the hair damage level, the pre-bleaching tone, and the target post-dyeing tone and acquires bleaching conditions including any of the name of a bleaching agent, the concentration of an oxidant during bleaching, the leaving time during bleaching, and the number of times of bleaching.

Description

High-tone hair color assistance program, system, method, and machine learning model

[0001] The present invention relates to a high-tone hair color assistance program, a high-tone hair color assistance system, a high-tone hair color method, and a high-tone hair color assistance machine learning model.

[0002] Hair dyeing technology can be mentioned as one means for satisfying people's fashion needs. Hair dyeing technology enables arrangement of hair color according to individual requests. Light hair color and the technology for achieving light hair color are called high-tone color, and there is much demand for this. The tone herein is an indicator of brightness in hair color.

[0003] However, there is a problem that even if a hairdresser wants to incorporate this high-tone color into their own beauty salon, they cannot introduce it. Except for some famous salons and persons who hail from such salons, the option of "including high-tone coloring in the service menu" cannot be obtained.

[0004] One reason for this is that experience in high-tone color (technology) cannot be obtained or is insufficient. This is because opportunities to acquire high-tone color (technology) are limited. Opportunities to come into contact with the latest trends and technologies are concentrated in urban areas, and even when such opportunities exist, high costs are required for acquiring practical skills through means such as attending seminars.

[0005] Furthermore, there are other underlying problems: there are no practical teaching materials for improving hair coloring technology (hair dyeing technology) on a daily basis, and even when attempting to recruit an excellent hairdresser capable of instructing high-tone coloring, the person's skills cannot be objectively grasped.

[0006] In addition, there are other problems in the hair salon industry: efficiency improvement through technology has not progressed, there is a large disparity in technology and information between urban and rural areas (as mentioned above), and the turnover rate is high due to factors such as long non-working hours, which prevents technology from taking root.

[0007] Patent Document 1 discloses a system that displays image information that allows users to grasp the color and / or brightness of hair after it has been dyed with each of a plurality of first hair coloring agents capable of dyeing hair to different colors and / or brightness, corresponding to information that can identify the color and / or brightness of hair before dyeing; a system that facilitates the calculation of the amount of each substance when mixing different substances; and a system that notifies that information that allows users to grasp the color and / or brightness of hair after it has been dyed has not been stored for a given hair coloring agent.

[0008] Japanese Patent Publication No. 2024-077011

[0009] Achieving a high-tone hair color requires sufficient bleaching (decolorization). If the bleaching is insufficient, the hair will remain dark even after dyeing, and the high-tone color will not be achieved. However, the bleaching technique required to achieve a high-tone color largely depends on the skill of the hairdresser, which is one of the reasons why mastering high-tone coloring is difficult.

[0010] Patent document 1 does not describe any consideration of this bleaching stage.

[0011] The problem we aim to solve is the lack of programs or systems that propose and provide users, such as hairdressers, with the necessary bleaching conditions before dyeing, which are required to achieve a high-tone color after dyeing. Another problem we aim to solve is the lack of a system to evaluate the results after application, such as the uniformity of color after applying the hair dye, and the skill level of the dyeing procedure.

[0012] The most important feature of this invention is a bleach condition provision program and a function that displays the bleach conditions using the bleach condition provision program.

[0013] The present invention has been made in view of the above problems, and employs, for example, the following means. That is, it provides a high-tone color assistance program characterized by making a computer function as: a hair damage level acquisition means for acquiring the hair damage level of a person who wishes to have their hair dyed before bleaching; a pre-bleach tone acquisition means for acquiring a pre-bleach tone, which is a numerical value indicating the brightness of the person's hair before bleaching; a target post-bleach tone acquisition means for acquiring a target post-bleach tone, which is a numerical value indicating the target brightness of the hair color desired by the person who wishes to have their hair dyed; and a bleach condition acquisition means for acquiring bleach conditions, which include the name of the bleaching agent, the concentration of the oxidizing agent during bleaching, the waiting time during bleaching, or the number of bleaching sessions, using the hair damage level, the pre-bleach tone, and the target post-bleach tone as input.

[0014] The high-tone color support program of the present invention has the advantage of making it easier for hairdressers to learn high-tone coloring (technique) by displaying the bleaching conditions before dyeing according to the hair of the person who wishes to dye their hair. Furthermore, the high-tone color support program of the present invention enables evaluation of post-application conditions, such as the uniformity of color after application of the hair dye, making it possible to evaluate the skill level of the hair dyeing procedure.

[0015] This is a diagram (network configuration diagram) showing an overview of the high-tone color support system 1. This is a diagram of the head image acquisition screen (before bleaching). This is a diagram of the post-dyeing image display screen. This is a diagram of the bleaching condition display screen. This is a diagram of the hair dyeing condition display screen. This is a diagram of the hair dyeing technique evaluation display screen. This is a flowchart of the bleaching condition provision process. This is a diagram of the judgment conditions for outputting bleaching conditions. This is a flowchart of the hair dyeing condition provision process. This is a flowchart of the hair dyeing technique evaluation provision process. This is a hardware configuration diagram of server 10. This is a hardware configuration diagram of terminal 20.

[0016] Embodiments of the present invention will be described with reference to the drawings. In the following embodiments, the same or corresponding parts are denoted by the same reference numerals, and their descriptions may be omitted as appropriate. Furthermore, the drawings used below are for the purpose of explaining these embodiments and may differ from the actual device configuration, user interface (UI), data configuration, etc.

[0017] (Outline of Embodiment) An overview of this embodiment will be described using Figure 1. Figure 1 is a diagram (network configuration diagram) showing an overview of the system that performs processing using the high-tone color support program P1 of this embodiment (hereinafter referred to as "high-tone color support system 1"). A user such as a beautician acquires an image of the head of a person who wants to have their hair dyed before bleaching (especially an image of the back of the head) using the imaging device (camera) of terminal 20, and transmits the bleached frontal image and a post-dye image that reflects the wishes of the person who wants to have their hair dyed to server 10. Server 10, which is equipped with the high-tone color support program P1, is equipped with various programs including a machine learning model (such as the bleach condition provision program P24). Server 10 acquires the bleached frontal image and post-dye image of the person who wants to have their hair dyed, and displays the bleach conditions. The bleach conditions include, for example, the name of the bleaching agent, the concentration of the oxidizing agent during bleaching, and the waiting time (after applying the bleaching agent). The user performs bleaching according to these bleach conditions. Furthermore, server 10 acquires the bleached frontal image and post-dye image mentioned above, and displays the dyeing conditions. Dyeing conditions include, for example, the type of dye and the dyeing time. With the high-tone color support system 1 of this embodiment, even users with little experience in hair dyeing can learn the bleaching and dyeing conditions necessary to obtain high-tone colors. Furthermore, by having the user take an image of the back of the head of a person who wishes to have their hair dyed with the terminal 20 and send it to the server 10, it is possible to obtain an evaluation of the hair dyeing technique related to the bleaching and dyeing procedures.

[0018] (Details of the Embodiment) The high-tone color support system 1 according to this embodiment will be described in detail below. The high-tone color support system 1 includes a computer (server 10) equipped with a high-tone color support program P1, and provides a system online that displays the conditions for high-tone coloring to users such as beauticians. In other words, the information processing by the high-tone color support program P1 in the high-tone color support system 1 is concretely realized using hardware resources. The following describes in order the following components that constitute the high-tone color support system 1: 1. User interface, 2. Program processing, 3. Data, and 4. Hardware configuration.

[0019] (Definition of Terms) Here, we define some terms. "High-tone (color)" refers to a light hair color. Here, tone is an index (or unit) and may be expressed as level. Generally, tones (levels) range from 1 to 20, with higher numbers indicating higher hair brightness. When referring to high-tone (color), the tone (level) is generally 12 or higher. In the following, the tone (level) and measurement conditions are based on the applicant's own index (sensory evaluation by a beautician with advanced expertise in hair dyeing technology), but since tone and level are affected by the lighting environment (sunlight, indoor lighting, etc.), they are measured according to predetermined standards (measurement conditions, etc.). These standards may conform to the standards of a designated organization. For example, the NPO Japan Hair Color Association provides a level scale that divides hair brightness into 12 stages from 4 to 15. Hair tone (level) of 12 or higher according to the standards of the said organization may be considered high-tone. "User" refers to a person who obtains bleaching conditions, etc., using the high-tone color support system 1. For example, this includes hairdressers and beauty salons that provide beauty and service services, including hair dyeing. This applies to both natural persons and legal entities. "Color category" is a category of color, expressed as a numerical value and / or string. The string here refers to, for example, "red," "orange," "yellow," etc. In the following embodiments, each category is distinguished by one numerical value and one string, such as "1: red," "2: orange," etc. In the following embodiments, the color categories are stored in advance in the storage unit 14 of the server 10. Also, the number of color categories in the following embodiments is 12. "Color space data" is data that represents color. For example, it is a general term for data that represents color information as data such as feature quantities (vectors) or numerical values, such as RGB, HSV, CIELAB (LAB), CIELUV (LUV), or XYZ, by representing color in space. It is usually expressed in a three-dimensional color space. The color space is sometimes expressed as a color system. "RGB" represents each component R (Red), G (Green), and B (Blue) with a value from 0 to 255."HSV" expresses the hue component as a value from 0 to 360 degrees, and the saturation and value components as values ​​from 0 to 100%. "CIELAB (LAB)" is a color space standardized by the International Commission on Illumination (CIE) (CIE1976 Lab color space) and adopted in Japan by JIS (JIS Z 8781-4). It expresses the value component (L star) as a value from 0 to 100, and the chromaticity, which indicates hue and saturation, as numerical values ​​(a star, b star). "CIELUV (LUV)" is a color space standardized by the International Commission on Illumination (CIE) (CIE1976 Luv color space). It has parameters for lightness (L star), redness / greenness (u star), and yellowness / blueness (v star). "XYZ" is a color space standardized by the International Commission on Illumination (CIE) (CIE1931XYZ color space). It has parameters X, Y, and Z. x + y + z = 1 (100%). "Bleaching agent" is a chemical used to decolorize hair. Bleaching agents may also be called hair bleach, hair lightener, or decolorizing agent. Bleaching agents generally contain alkaline substances such as ammonia or monoethanolamine. Bleaching agents may also contain persulfates or percarbonates to enhance their decolorizing power. It is common to combine bleaching agents with oxidizing agents (described below) during bleaching, but bleaching agents may also contain oxidizing agents (however, in this case, the bleaching power is not high). "Oxidizing agent" specifically refers to hydrogen peroxide itself or chemicals containing hydrogen peroxide. Oxidizing agents are sometimes referred to as "oxy". In the following, when we refer to the oxidizing agent concentration (oxy concentration), this refers to the concentration of hydrogen peroxide, which functions as an oxidizing agent. In Japan, concentrations of 3.0%, 4.5%, and 6.0% are common. The effect of the oxidizing agent (oxy) is to decompose the melanin contained in the hair, thereby bleaching it. During bleaching, alkaline substances and oxidizing agents generally act on the hair. The alkaline substance swells the hair, and the oxidizing agent decomposes the melanin and other substances contained in the hair, thereby bleaching it.In some cases, oxidation accelerators such as persulfates and percarbonates are also used. When bleaching, the bleaching agent containing alkali is sometimes referred to as the "first agent," the oxidizing agent (oxygen) as the "second agent," and the agent (bleaching agent) containing persulfates or percarbonates as the "third agent." In this embodiment, "oxidizing agent concentration (oxygen concentration)" refers to the concentration or (corresponding) amount of the oxidizing agent (oxygen) added as the second agent. This is also true for hair dyes, which will be discussed in the next section, but the oxidizing agent (concentration) used in bleaching and the oxidizing agent (concentration) used in hair dyeing have some differences in their uses. For example, the oxidizing agent (oxygen) used in hair dyeing plays a role in polymerizing the dye. This is because the agents used as bleaching agents and the agents used as hair dyes have different purposes and components. "Hair dyeing" refers to coloring the hair (making the hair color), and is sometimes referred to as color, hair color, coloring, or hair coloring. Furthermore, "hair dyes" used for coloring hair are also called colorants, hair colorants, coloring agents, or hair coloring agents. Sometimes, hair dyes are simply referred to as color or hair color. Hair dyes contain coloring agents (color developers) such as dyes, and in this respect, they differ from bleaching agents, for example. Hair dyes may or may not contain alkaline substances. Also, as mentioned above, oxidizing agents (oxidizers) may be used when dyeing hair. "People who want to dye their hair" refers to anyone who wants to dye their hair, but this definition is not limited to specific individuals. For example, it does not have to be limited to people who actually dye their hair; it could also refer to people who are simply interested in dyeing their hair. In a broad sense, people who want to dye their hair include customers of the aforementioned users (hairdressers, etc.) and people who are to be dyed, and also includes, for example, another hairdresser who dyes another hairdresser's hair for practice. While there is no need to define a clear boundary between the "root of the hair" and the "tip of the hair," for the sake of understanding, the root will be defined as approximately 5 cm from the hair follicle at one end of the hair, and the tip as approximately 5 cm from the other end of the hair. The "target hair color after dyeing" is the desired hair color after dyeing for the person requesting the hair dyeing, and the "actual hair color after dyeing" is the actual hair color after dyeing for the person requesting the hair dyeing. The evaluation items, "color reproduction accuracy," "gradation," "uniformity of application," and "coloring evaluation," will be explained in the section on hair dyeing technique evaluation.

[0020] In the following, the directions of up, down, front, back, left, and right are based on a person standing upright. That is, the head is up and the feet are down. It can also be said that the top of the head is up and the chin is down. The up and down direction is also called the vertical direction. The direction along the up and down direction of the hair is sometimes called the vertical direction. Furthermore, the direction that the person standing upright is facing is forward, that is, the front is the belly side and the back is the rear. In addition, from the perspective of the person standing upright, the right ear side is the right and the left ear side is the left. The front, back, left, and right directions are also called the horizontal direction. Furthermore, the left and right direction of the hair or the horizontal direction is sometimes called the lateral direction.

[0021] In the following, when the term "○○ process" is used, it means that the computer's processor executes a process based on the "○○" program stored in the program storage unit. The same word will be used in place of "○○" throughout this paragraph. That is, the "○○" program is a program that enables the computer to function as a "○○" means by executing the "○○" process. In this case, the control unit equipped with the processor also functions as a "○○" unit (or "○○" device). In this case, the "○○" unit executes the "○○" process based on the "○○" program. The "○○ process" may also be expressed as a "○○" method that includes multiple steps (or procedures) in chronological order.

[0022] For example, the high-tone color assistance program P1 is a program that causes the computer to function as a high-tone color assistance means by executing high-tone color assistance processing. In this case, the control unit 12 of the computer equipped with the processor 122 functions as a high-tone color assistance unit (or high-tone color assistance device).

[0023] In the high-tone color assistance system 1, each terminal (computer), such as the server 10 and terminal 20, is equipped with a processor. However, when simply referring to a processor, it refers to the processor that performs processing using the high-tone color assistance program P1, and in this embodiment, it refers to the processor 122 of the server 10.

[0024] In the following, for simplicity, the action of "the processor 122 of server 10 receiving a request from a terminal (terminal 20) and returning data to be displayed in the terminal's browser" may be described as "the processor 122 displays (causes) the data to be displayed (causes) the data to be displayed in the terminal's (terminal 20) browser" or "the processor 122 displays (causes) the data to be displayed." Similarly, the action of "the processor 122 of server 10 causing the data to be stored in the data storage unit 14b of the storage unit 14" may be described as "the processor stores (causes) the data to be stored." In addition, a machine learning model may be simply referred to as a "model."

[0025] In the following, letters (aa, x.x%, XX minutes, etc.) are used instead of various numerical values, such as the mixing ratio of chemicals (bleach or hair dye), the concentration of oxidizing agents, or the processing time after application. However, these do not restrict the numerical values ​​in any way. For example, aa and bb do not necessarily indicate that they are always different numbers (i.e., aa and bb may be the same number). Also, the number of letters does not limit the number of digits (or significant figures) of the number. For example, aa does not necessarily represent a two-digit number; it may represent a single digit or a number with decimal points. This is also true when substituting non-numeric characters, such as strings of characters. For example, "xxxxxx.jpg" indicates that the extension is ".jpg," but the file name is arbitrary. In the following tables, "..." means that data, rows, or columns are omitted.

[0026] 1. User Interface (UI) First, the interface that the high-tone color assistance system 1 of this embodiment displays on the terminal 20 will be explained using diagrams. The interface described below is a simplified version of the one that the processor 122 displays on the browser of the terminal 20.

[0027] Furthermore, only icons related to functions necessary for explanation will be displayed, and other publicly known icons will be omitted. For example, the back button for returning to the previously displayed page has been omitted.

[0028] Figure 2 shows the head image acquisition screen (before bleaching). The processor 222 of terminal 20 displays the image acquired in real time by the imaging device of terminal 20 on the display unit 284a, and also displays a head guide. The user positions the head of the person who wants their hair dyed (in this embodiment, the back of the head including the top of the head) to match this head guide and takes a photograph. The processor 222 of terminal 20 transmits the bleached forehead image acquired by this photograph to the server 10.

[0029] As shown in Figure 2, the head image acquisition screen (before bleaching) includes a real-time image display unit UI-11, a head guide UI-111 (left / right guide UI-111a and front / back guide UI-111b), and a capture button UI-12.

[0030] The real-time image display unit UI-11 displays images acquired in real time by the imaging device (for example, the camera of the smartphone if the terminal 20 is a smartphone) equipped in the terminal 20.

[0031] The head guide UI-111 is a guide displayed on the real-time image display unit UI-11 for aligning the head position. In this embodiment, the head guide UI-111 displays the position and orientation of the ears. The head guide UI-111 also includes left / right guides UI-111a and front / back guides UI-111b (white dotted lines in Figure 2).

[0032] The left-right guide UI-111a is a line indicating the left-right direction of the head, passing through the left and right sides of the head, for example, the right ear, the top of the head, and the left ear. The front-back guide UI-111b is a line indicating the front-back direction of the head, passing through the front-back sides of the head, for example, the forehead, the top of the head, and the back of the neck.

[0033] The user adjusts the position of the imaging device on terminal 20 so that the back of the head of the person requesting hair dyeing, as displayed on the real-time image display unit UI-11, matches the size of the head guide UI-111, and so that the orientation of the person's head matches the left / right guide UI-111a and the front / back guide UI-111b.

[0034] In particular, the head guide UI-111 displays the position and orientation of the ears, making it easier for users to understand what angle of view to use when taking a picture. Similarly, the presence of left / right guide UI-111a and front / back guide UI-111b makes it easier for users to understand what angle of view to use when photographing the top of a person's head who wants their hair dyed.

[0035] The capture button UI-12 is a button for the user to take a picture. That is, the processor 222 acquires the image (head image) displayed on the real-time image display unit UI-11 at the moment the button is pressed.

[0036] After capturing the image, the processor displays a confirmation screen asking whether it is OK to send the head image to the server 10. Once the user confirms (for example, by pressing the "OK" button), the processor sends the head image to the server 10. In other words, the server 10 acquires the head image through the terminal 20.

[0037] Figure 3 shows the post-dyeing image display screen. As shown in Figure 3, the processor 222 of terminal 20 displays the post-dyeing image on the post-dyeing image display unit UI-21. The "post-dyeing image" is a head image showing the image after dyeing (particularly the image related to color and tone). For example, a person who wants to dye their hair looks at this post-dyeing image and decides on their desired color (coloring). Note that the "post-dyeing image" is different from the "completed dyeing image" on the dyeing conditions display screen, which will be described later.

[0038] In this embodiment, the post-dyeing image is either an image included in a catalog pre-stored in the storage unit of the server 10 or terminal 20, or an image presented by the person requesting the hair dyeing.

[0039] The images presented by those requesting hair dyeing originate from, for example, images on a device (such as a smartphone) owned by those requesting hair dyeing, and are ultimately obtained by the server 10 by transferring the images (or similar images) stored on the device owned by those requesting hair dyeing to the terminal 20.

[0040] Here, the method for transferring the post-dyeing image from a terminal owned by a person desiring hair coloring or the like to the terminal 20 is not particularly limited. For example, the method may be a method using online transmission and reception, a method using a storage medium, or a method of capturing, with the terminal 20, an image displayed on the terminal of the person desiring hair coloring or the like.

[0041] Further, the catalog is, for example, an image group including reference images related to hair colors, such as model images of various hair colors. The "person desiring hair coloring or the like" refers to, in addition to the person desiring hair coloring, for example, a related party (such as a friend) of the person desiring hair coloring, or a beautician.

[0042] The processor 222 transmits, to the server 10, the selection of a post-dyeing image (when selected from the catalog) or the post-dyeing image (when it is an image presented by the person desiring hair coloring or the like).

[0043] Note that, for example, when a catalog of images is stored in the storage unit 14 of the server 10, the server 10 can acquire information (such as an image number) linked to the images in the catalog. This is because the server 10 can acquire the post-dyeing image from the information linked to the image.

[0044] Although details will be described later, the processor 122 recognizes the head included in the acquired post-dyeing image, and acquires a "target post-dyeing tone", which is a numerical value indicating the (target) brightness of the post-dyeing hair color desired by the person desiring hair coloring, and a "target post-dyeing color category", which is the (target) color category of the post-dyeing hair color desired by the person desiring hair coloring. As described above, the post-dyeing hair color desired by the person desiring hair coloring is referred to as the "target post-dyeing hair color".

[0045] More specifically, the processor 122 detects a head from a selected image such as an image presented by the person desiring hair coloring or the like, extracts color space data from the central part of the detected head (head central part), and determines which of predefined color categories and tones the extracted data matches.

[0046] Note that in the present embodiment, the central part of the head refers to the center (and the surrounding area thereof) of a bounding box used for head detection. The designer of a machine learning model or the like can appropriately determine how large an area is to be set as the central part.

[0047] As shown in FIG. 3, the post-hair-dyeing image display screen includes a post-hair-dyeing image display unit UI-21 and an operation UI display unit UI-22.

[0048] The post-hair-dyeing image display unit UI-21 displays an image of the color after hair dyeing. That is, the processor 122 causes the post-hair-dyeing image display unit UI-21 to display an image included in a catalog or an image provided by a person desiring hair dyeing or the like.

[0049] The operation UI display unit UI-22 displays icons and the like for use by a user's operation. Examples thereof include a button for a user to select an image and a button for a user to determine an image. The processor 222 can display known operation icons used for image selection, determination, and the like on the operation UI display unit UI-22.

[0050] When the user performs an operation to determine an image, the processor 222 transmits the head image displayed on the post-hair-dyeing image display unit UI-21 to the server 10. The processing of the server 10 will be described later.

[0051] FIG. 4 is a diagram showing a bleaching condition display screen. The terminal 20 acquires bleaching conditions from the server 10 and displays the bleaching conditions on a display unit 284a of the terminal 20. As shown in FIG. 4, in the present embodiment, the terminal 20 displays the bleaching conditions and an image after completion of bleaching.

[0052] As shown in FIG. 4, the bleaching condition display screen includes a bleaching condition display unit UI-31 and a post-bleaching completion image display unit UI-32.

[0053] In the present embodiment, the processor 122 causes the bleaching condition display unit UI-31 to display bleaching conditions, namely, the name(s) of one or more bleaching agents, the mixing ratio of the bleaching agents when there are two or more bleaching agents, the oxidizing agent concentration during bleaching, and (a guideline for) the standing time after application of the bleaching agent.

[0054] For example, FIG. 4 shows the bleaching conditions for the first bleaching, and the bleaching condition display screen indicates that a predetermined amount (aa) of bleaching agent A and a predetermined amount (bb) of bleaching agent B are mixed as the bleaching agents.

[0055] Note that this quantity can be expressed as either weight (ratio) or mass (ratio). This also applies to the following hair dyes.

[0056] The bleaching conditions display screen shows the bleaching conditions for the roots and ends of the hair. Here, the specified oxidizing agent concentrations (x.x% and y.y%) and the waiting time (XX minutes and YY minutes) (after applying the bleach) are displayed for each. Note that the waiting time can also be called the bleaching time.

[0057] Furthermore, as shown in Figure 4, if it is necessary to change the bleaching conditions at the roots and ends of the hair, the processor 122 displays the bleaching conditions for the roots and ends of the hair, respectively.

[0058] The post-bleaching image display unit UI-32 displays an image of the system after the bleaching is complete (post-bleaching image). The post-bleaching image is data that has been pre-stored in the storage unit 14 of the server 10. The storage unit 14 stores the corresponding post-bleaching image for each bleaching condition, which will be described later.

[0059] Furthermore, if the bleaching conditions include multiple bleaching operations, the image (picture) after each bleaching operation of server 10 is displayed on the post-bleaching image display unit UI-32 (not shown). For details, please refer to the bleaching condition database D10.

[0060] The same applies if the number of bleaching sessions is two or more. The explanation of the bleaching condition display screen when the number of bleaching sessions is two or more is omitted.

[0061] Figure 5 shows the hair dyeing conditions display screen. As shown in Figure 5, terminal 20 acquires data from server 10 and displays the hair dyeing conditions and the completed hair dyeing image on display unit 284a. As mentioned above, the "post-dyeing image" on the post-dyeing image display screen and the "completed hair dyeing image" on the hair dyeing conditions display screen are different images.

[0062] As shown in Figure 5, the hair dyeing condition display screen includes a hair dyeing condition display unit UI-41 and a hair dyeing completion image display unit UI-42.

[0063] In this embodiment, the hair dyeing condition display unit UI-41 displays the hair dyeing conditions (color conditions), namely the name of the hair dye, the amount of hair dye (or mixing ratio), and the approximate waiting time for the hair dye.

[0064] For example, in Figure 5, the hair dyeing conditions display screen shows the amount of hair dye to be used and the waiting time (after applying the hair dye) for both the roots and ends of the hair.

[0065] Specifically, the processor 122 displays on the hair dyeing condition display screen the mixing ratio of the hair dye to be applied to the roots of the hair (hair dye A: hair dye B = cc:dd) and the waiting time after application, as well as the mixing ratio of the hair dye to be applied to the ends of the hair (hair dye A: hair dye C = ee:ff) and the waiting time after application.

[0066] Although not shown in Figure 5, if an oxidizing agent (oxygen) is used separately from the hair dye, the processor 122 displays the concentration of the oxidizing agent during hair dyeing. In this case, the processor 122 also displays the mixing ratio of the hair dye and the oxidizing agent (oxygen).

[0067] Furthermore, as shown in Figure 5, if it is necessary to change the dyeing conditions at the roots and tips of the hair, the processor 122 displays the dyeing conditions for the roots and tips of the hair, respectively.

[0068] The hair dyeing completion image display unit UI-42 displays an image of the hair after dyeing is complete (hair dyeing completion image). The image of the hair after dyeing is data that is stored in advance in the storage unit 14 of the server 10. The storage unit 14 stores the corresponding hair dyeing completion image for each hair dyeing condition, which will be described later.

[0069] Figure 6 shows the hair dyeing technique evaluation display screen. After the procedure (hair dyeing), the user takes a picture of the back of their head with the terminal 20 and sends it to the server 10. The terminal 20 then obtains a hair dyeing technique evaluation from the server 10 and displays it on the display unit 284a.

[0070] Terminal 20 acquires data from server 10 and displays the coloring evaluation, color reproduction accuracy, gradation, and uniformity of application as part of the hair dyeing technique evaluation. As shown in Figure 6, the coloring evaluation displayed in the upper row is 3.0, and the color reproduction accuracy, gradation, and uniformity of application displayed in the lower row are 2.0, 3.0, and 4.0 from left to right.

[0071] Color reproduction accuracy, gradation, and uniformity of color application are evaluated individually, while the coloring evaluation is an overall evaluation that takes these individual evaluations into account. Terminal 20 can also acquire data from server 10 and display the analysis results.

[0072] The processor 222 displays a head image acquisition screen (after hair dyeing) for capturing an image of the back of the head after hair dyeing. However, the function of this screen is the same as that described in the head image acquisition screen (before bleaching) (Figure 2), so the explanation is omitted.

[0073] As shown in Figure 6, the hair dyeing technique evaluation display screen includes an evaluation display unit UI-51 and an analysis result link unit UI-52.

[0074] The evaluation display unit UI-51 displays the coloring evaluation, color reproduction accuracy, gradation, and uniformity of the paint application. The evaluation (scoring) method will be described later.

[0075] The analysis results link UI-52 is a link to display the analysis results display screen (not shown). The analysis results display screen is a screen for the user to check various data related to the evaluation of hair dyeing techniques.

[0076] Specifically, the processor 122 displays the bleached frontal image, the post-dyeing image, the bleaching conditions, the image after bleaching is complete, the dyeing conditions, the image after dyeing is complete, and the dyed back of the head image on the analysis results display screen (displayed by scrolling through the images by swiping). From the analysis results display screen, the user can visually confirm the head image before the treatment, the target image, the treatment recipe (bleaching conditions and dyeing conditions), and the actual finished result of the treatment.

[0077] With the above configuration, users can input an image of the client's hair before the coloring treatment and a target image of the hair after coloring to confirm the bleaching and coloring conditions. Furthermore, the processor 122 displays a completed image after bleaching, allowing users to see the actual bleached result and confirm whether the bleaching was performed as planned. In this way, the high-tone color support system 1 can present bleaching conditions in particular, allowing users to perform the bleaching, which is essential for high-tone coloring, with confidence.

[0078] 2. Program Processing <High-tone color support processing> The program processing performed in the high-tone color support system 1 of this embodiment will be described below.

[0079] In this embodiment, the processor 122 performs high-tone color support processing based on the high-tone color support program P1. The high-tone color support program P1 includes at least a bleach condition provision program P12, a hair dye condition provision program P14, and a hair dye technique evaluation provision program P16, and the processor 122 performs bleach condition provision processing, hair dye condition provision processing, and hair dye technique evaluation provision processing based on each of these programs.

[0080] <2-1. Bleach Condition Provisioning Process> The processor 122 performs bleach condition provisioning processing based on the bleach condition provisioning program P12. That is, the bleach condition provisioning program P12 makes the computer function as a bleach condition provisioning means by executing the bleach condition provisioning processing by the processor 122.

[0081] In the bleaching condition provision process, the processor 122 performs the following: a questionnaire response acquisition process to acquire responses to a questionnaire including a question about the hair damage level of the person requesting hair dyeing before bleaching; a hair damage level acquisition process to acquire the hair damage level of the person requesting hair dyeing before bleaching; a bleached forehead image acquisition process to acquire a bleached forehead image, which is an image of the head of the person requesting hair dyeing before bleaching; a natural hair exposure determination process to determine whether or not the natural hair is visible from the bleached forehead image; a pre-bleach tone acquisition process to acquire a pre-bleach tone, which is a numerical value indicating the brightness of the hair before bleaching from the bleached forehead image; a target post-dye tone acquisition process to acquire a target post-dye tone, which is a numerical value indicating the target brightness of the hair color after dyeing desired by the person requesting hair dyeing; a bleaching condition acquisition process to acquire bleaching conditions including the name of the bleaching agent, the concentration of the oxidizing agent during bleaching, the waiting time during bleaching, and the number of bleaching sessions, using the hair damage level, the pre-bleach tone, and the target post-dye tone as inputs; and A bleach condition display process is executed, which displays the aforementioned bleach conditions on the display unit 284a of the user's terminal (terminal 20).

[0082] Here, the bleach condition acquisition process is performed by the processor 122 executing the bleach condition provision program P24.

[0083] Furthermore, in this embodiment, the bleach condition acquisition process, which takes the hair damage level, the tone before bleaching, and the target tone after dyeing as inputs to acquire bleach conditions including the name of the bleaching agent, the oxidizing agent concentration during bleaching, the waiting time during bleaching, and the number of bleaching sessions, is also a bleach condition acquisition means, which takes the hair damage level, the presence or absence of natural hair exposure, the tone before bleaching, and the target tone after dyeing as inputs to acquire bleach conditions including the name of the bleaching agent, the root oxidizing agent concentration (the concentration of the oxidizing agent at the root of the hair during bleaching), the tip oxidizing agent concentration (the concentration of the oxidizing agent at the tip of the hair during bleaching), the root waiting time (the waiting time at the root of the hair during bleaching), the tip waiting time (the waiting time at the tip of the hair during bleaching), and the number of bleaching sessions.

[0084] In this embodiment, the processor 122 performs a pre-bleach color category acquisition process to acquire a pre-bleach color category, which is the color category of the head in the bleached forehead image, and a target post-dye color category acquisition process to acquire a target post-dye color category, which is the color category of the hair color after dyeing desired by the person requesting the hair dyeing.

[0085] Processor 122 uses this pre-bleach color category for evaluating hair dyeing techniques, as described later. Processor 122 also uses this target post-dye color category in the hair dyeing condition provision process, as described later.

[0086] Figure 7 is a flowchart of the bleach condition provision process. The processor 122 starts the bleach condition provision process after receiving input such as survey information from the user after the application program is launched. In the following diagrams, steps are abbreviated as "S". For example, step 1 is represented as S1.

[0087] Processor 122 obtains the hair damage level of the person requesting hair dyeing before bleaching (Step 1: Hair damage level acquisition process).

[0088] Hair damage refers to damage to the hair, such as loss of firmness, body, shine, or elasticity, loss of moisture or protein from within the hair, peeling of the cuticle, or the presence of split ends. This damage can be evaluated using equipment or by sensory evaluation. Hair damage information refers to information about this hair damage (whether the hair is damaged or not), and includes, for example, the loss of moisture or protein from within the hair, the presence or absence of split ends, and information combining these.

[0089] In this embodiment, hair damage information is obtained in the form of responses to a questionnaire displayed on the display unit 284a of the terminal 20 by the processor 122. The questionnaire asks, "Is your hair damaged?", and the respondent answers either "yes" or "no". The terminal 20 receives this response and transmits it to the server 10.

[0090] Here, the process of obtaining responses from users such as those who wish to dye their hair or hairdressers to a questionnaire that includes a question about the level of hair damage before bleaching is referred to as the questionnaire response acquisition process.

[0091] The survey results are not limited to a simple yes / no choice; they may also be based on a multi-level scale, such as a five-point scale. This survey can be completed by individuals who wish to have their hair dyed, or by users such as hairdressers who assess the condition of a client's hair.

[0092] The reason for obtaining information about hair damage is that the bleaching conditions need to be adjusted according to the state of hair damage. In other words, bleaching damaged hair can severely damage it, so if the hair damage is significant, the bleaching conditions are generally made gentler. Therefore, obtaining this hair damage information is essential when bleaching.

[0093] As shown in Figure 7, the processor 122 acquires a bleached forehead image, which is an image of the head of a person who wants to dye their hair before bleaching (Step 2: Bleached Forehead Image Acquisition Process).

[0094] The user takes a photograph of the head of the person who wants their hair dyed (in this embodiment, the back of the head including the crown (e.g., the whorl)) and sends it to the server 10. The processor 122 acquires this image information.

[0095] In this embodiment, the head determination model P21 is used to detect and acquire various head images (bleached forehead image, dyed hair back image) from images taken by the imaging device of the terminal 20. Details of the head determination model P21 will be described later, but the head determination model P21 is a machine learning model that determines the head from images acquired by the imaging device using object recognition technology.

[0096] The processor 122 also determines whether or not natural hair is visible from the acquired bleached forehead image (Step 3: Natural hair visibility determination process).

[0097] In this embodiment, determining the appearance of natural hair means determining whether the roots of the hair at the crown of the head are dyed or not. For example, if the original hair color is black, the state in which undyed natural hair appears at the roots of the hair due to hair growth is generally called the "pudding state" (the term "pudding" comes from the dark colored caramel on top).

[0098] For example, in the case of hair, 10% might be black and 90% dyed, but here, regardless of this ratio, the term "pudding state" refers to a situation where the natural hair color and the dyed color coexist.

[0099] In this embodiment, the presence or absence of natural hair is determined by the natural hair manifestation determination model P22. The details of the natural hair manifestation determination model P22 will be described later, but the natural hair manifestation determination model P22 makes a binary determination as to whether or not natural hair is manifested after dyeing, that is, whether or not there is an undyed portion at the root of the hair. The determination result obtained here will be called the "manifestation determination result".

[0100] The manifestation determination result is not limited to binary data such as whether or not the hair is manifested, but may also be based on the ratio of the black area to the dyed hair area (e.g., black area to dyed hair area = 1:9, 2:8).

[0101] Processor 122 obtains the pre-bleach tone, which is a numerical value indicating the brightness of the hair before bleaching, from the bleached forehead image acquired in step 2 (step 4: pre-bleach tone acquisition process). The pre-bleach tone is the tone of the hair portion in the bleached forehead image.

[0102] In this embodiment, a hair tone color determination model P23 is used to obtain the hair color tone of a person who wishes to dye their hair before bleaching. Details of the hair tone color determination model P23 will be described later, but in this step, the hair tone color determination model P23 outputs the hair color tone (level) of the person who wishes to dye their hair before bleaching as an integer.

[0103] In this embodiment, the processor 122 also acquires the pre-bleach color category, which is the color category of the head in the bleached frontal image (pre-bleach color category acquisition process).

[0104] As shown in Figure 7, the processor 122 acquires an image of the hair after dyeing (post-dyeing image acquisition process).

[0105] As described above, the post-dyeing image is either an image included in a catalog pre-stored in the storage unit of server 10 or terminal 20, or an image presented by the person requesting the hair dyeing. The processor 122 obtains the post-dyeing image by selecting an image from the catalog or by acquiring an image presented by the person requesting the hair dyeing.

[0106] In this embodiment, if the post-dyeing image is an image presented by the person requesting hair dyeing, the processor 122 recognizes the head from the post-dyeing image and obtains the color category and tone of the head (center) (referred to as "post-dyeing image analysis processing"). In the post-dyeing image analysis processing, the processor 122 uses the hair tone color determination model P23 to obtain the color category and tone of the image presented by the person requesting hair dyeing.

[0107] In this embodiment, the catalog images stored in the memory unit 14 are pre-associated with color categories and tones. This eliminates the need for post-dyeing image analysis processing, thus shortening processing time.

[0108] Processor 122 obtains a target post-dye tone, which is a numerical value indicating the target brightness of the hair after dyeing (after bleaching and then dyeing), through post-dye image analysis processing (Step 6: Target post-dye tone acquisition process).

[0109] The target hair color tone after dyeing is the desired hair color tone for the person requesting the hair dyeing. In the case of the catalog mentioned above, it is the hair color tone in the catalog image, and in the case of the post-dyeing image mentioned above, it is the hair color tone in that image.

[0110] In this embodiment, a hair tone color determination model P23 is used to obtain the target post-dye tone. In this step, the hair tone color determination model P23 outputs the tone (level) of the post-dye hair color desired by the person requesting the hair dyeing, i.e., the target post-dye tone, as an integer.

[0111] The processor 122 of this embodiment obtains the target post-dye color category, which is the color category of the hair color desired by the person who wants to have their hair dyed, through post-dye image analysis processing (referred to as "target post-dye color category acquisition processing").

[0112] In this step, the hair tone color determination model P23 outputs the color category of the desired post-dye hair color for the person requesting hair dyeing, i.e., the target post-dye color category. The processor 122 uses the target post-dye color category to determine the dyeing conditions, which will be described later.

[0113] To summarize, the post-dyeing image analysis process includes the process of acquiring the target post-dyeing tone and the process of acquiring the target post-dyeing color category.

[0114] As shown in Figure 7, the processor 122 obtains bleaching conditions (name of the bleaching agent, concentration of the oxidizing agent during bleaching, standing time during bleaching, and number of bleaching cycles) using the bleaching condition provision program (Step 7: Bleaching Condition Acquisition Process).

[0115] In this embodiment, the bleaching condition provision program P24 is used to obtain the bleaching conditions. Details of the bleaching condition provision program P24 will be described later, but the bleaching condition provision program P24 outputs the bleaching conditions necessary to achieve the desired post-dye hair color of the person who wants to have their hair dyed.

[0116] The processor 122 then displays the aforementioned bleaching conditions on the display unit 284a of the user's terminal (terminal 20) (Step 8: Bleaching Condition Display Process). For a specific example, please refer to the bleaching condition display screen (Figure 4).

[0117] As mentioned above, if the hair damage level is high (when the damage level is evaluated on a multi-level scale and exceeds a certain value), the procedure carries risks. Therefore, the processor 122 displays a pop-up on the display unit 284a of the terminal 20 to prompt the person requesting hair dyeing for confirmation. In other words, if the damage level is above a certain level (i.e., there is damage), the processor 122 displays a warning screen on the display unit 284a of the terminal 20 (referred to as the "warning screen display process").

[0118] (Description of each program) Here, we will explain each of the above-mentioned programs (head determination model P21, natural hair exposure determination model P22, hair tone color determination model P23, and bleach condition provision program P24). Note that in the following, the processor used to train the machine learning models is not necessarily processor 122. A personal computer (PC) suitable for training machine learning models can be used as appropriate.

[0119] The head identification model P21 is a machine learning model that uses object recognition technology to identify heads from images acquired by imaging devices, etc. In other words, it is an object detection type machine learning model that detects a person's head from an image that includes a person's head.

[0120] The head recognition model P21 learns using images containing human heads and the information (annotation data) attached to those images as training data. For example, a model creator creates a machine learning model by indicating (annotating) the region that is judged to be a head (usually a rectangular area called a bounding box) in an image containing a human head. In other words, the model creator teaches the computer the target object (in this case, a human head) and the position of the target object on the image. This target object and its position constitute the annotation data.

[0121] Through this learning process, the head recognition model P21 outputs (detects) the region (location) of a human head if it is present in the image, and displays it. In addition, it may output a confidence score that quantifies the confidence level of the output result.

[0122] For example, the head identification model P21 takes a bleached forehead image as input, identifies the head in the bleached forehead image, and outputs the result. The head identification model P21 is also used in the hair dyeing technology evaluation provision program P16, which will be described later. This is because a head image is required when sending a dyed back of the head image to the server 10 during the hair dyeing technology evaluation provision process.

[0123] The learning process in this embodiment will now be described in detail. Supervised learning is applied to the training of the head recognition model P21, and the training data consists of 400 images for training and 80 images for testing, totaling 480 images, along with corresponding annotation data. A bounding box is fitted to the head, and the coordinates of the rectangle are described in a format suitable for YOLO (You Only Look Once) and used for training.

[0124] In head detection, to prevent the model from learning that heads with different hair colors are different objects, all training data is converted to grayscale before the model is trained. Converting to grayscale enables more efficient model training.

[0125] In addition, when training the head recognition model P21 in this embodiment, images of the training data that have been horizontally flipped are used.

[0126] As described above, YOLO can be suitably used as an object detection algorithm, for example. In this embodiment, YOLOv8 from Ultralytics® is used.

[0127] In addition, other object detection algorithms such as R-CNN (Region Based Convolutional Neural Networks), FPN (Feature Pyramid Networks), SSD (Single Shot Multibox Detector), or cascade classifiers (OpenCV) may be used. Publicly known object detection techniques may be used as appropriate.

[0128] Furthermore, publicly available tools can be used as appropriate for annotation.

[0129] The natural hair exposure determination model P22 is a machine learning model that uses an image including the head (especially the crown) of a person with dyed hair to determine, in a binary choice, whether or not natural hair is visible, that is, whether or not there is undyed hair at the hair roots. In this embodiment, YOLOv8 from Ultralytics® is used.

[0130] The natural hair manifestation detection model P22 is trained using images that include a person's head (especially the top of the head) where natural hair is not visible, images that include a person's head (especially the top of the head) with dyed hair where natural hair is visible, and label data for each of these images as training data. For example, images where natural hair is not visible are labeled "0", and images where natural hair is visible are labeled "1".

[0131] As a result, the natural hair visibility determination model P22 takes images from imaging devices (especially images including human heads) as input and outputs whether or not natural hair is visible in the image (binary data). In other words, this machine learning model assigns a single class (label) to the entire image.

[0132] For example, the natural hair visibility determination model P22 takes a bleached forehead image as input and outputs whether or not natural hair is visible.

[0133] The learning process in this embodiment will now be explained in detail. The natural hair exposure detection model P22, like the head detection model P21, applies supervised learning. While head detection performs "object detection," that is, the process of calculating the coordinates of the target object in the image, the natural hair exposure detection model performs processing related to "image classification."

[0134] The image classification process involves determining whether the entire image is of the target object, and the model's output is an integer, not coordinates. This integer corresponds to the correct label of the pre-trained data, and in this model, two types of output are possible: "0: No visible hair" and "1: Visible hair".

[0135] The P22 model for determining natural hair appearance uses a total of 192 images as training data.

[0136] For example, YOLO (You Only Look Once) can be suitably used as such an image classification algorithm. In this embodiment, YOLOv8 from Ultralytics® is used, but it is not limited to this, and any known image classification algorithm can be used. Furthermore, any known image classification technology can be used as appropriate.

[0137] In this embodiment, an image of the top of the head is used because it is easy to detect natural hair. However, it is sufficient to determine that the natural hair color has appeared at the roots of the hair due to hair growth some time after dyeing, and the image detection location is not necessarily limited to the top of the head. For example, a photograph of the person who wants to dye their hair could be taken from the front, and detection could be performed on the roots of the bangs (however, in this case, a separate set of such bangs images would need to be prepared as training data).

[0138] The hair tone and color determination model P23 is a machine learning model that determines the tone (level) and color category of hair.

[0139] In this embodiment, the hair tone color determination model P23 is trained using the following data: an image of a person's head (head image) acquired by an imaging device, the tone (level) (integer value) of the hair color in the central part of the head, and the color category of the hair color in the central part of the head. As described above, in this embodiment, the central part of the head is the central part of the bounding box used for head detection, and the designer of the machine learning model can appropriately determine how large an area to consider as the central part.

[0140] As a result, the hair tone color determination model P23 takes a head image as input and outputs the tone (level) of the hair color of the head (related to the input) as an integer, and also outputs the color category of the hair color of the head (related to the input).

[0141] Specifically, the hair tone color determination model P23 outputs a pre-bleach tone, which is a numerical value indicating the brightness of the hair before bleaching, if it is performing the pre-bleach tone acquisition process (step 4), and outputs a target post-dye tone, which is a numerical value indicating the target brightness of the hair after bleaching and dyeing (after bleaching and then dyeing) if it is performing the target post-dye tone acquisition process (step 6).

[0142] In other words, the system may take an image of the bleached forehead of a person who wants their hair dyed as input and output the tone (level) of the hair color of that person (in the image) before bleaching as an integer (step 4), or it may take an image after dyeing as input and output the tone (level) of the hair color in that image after dyeing as an integer (step 6).

[0143] The learning process in this embodiment will now be explained in detail. Supervised learning is applied to the training of the hair tone color determination model P23. Images of human heads (back of the head) (especially images of various hair colors) and the tone and color category of the head in those images are used as training data. Here, the tone and color category of the head in those images are determined by a beautician with advanced expertise in hair dyeing techniques (sensory evaluation). For example, for a given head image, the color category is "red" and the tone is "12," and so on, with training data (ground truth data) assigned to both the color category and the tone.

[0144] On the other hand, the above-mentioned training images are also subjected to mechanical analysis. Specifically, the hair tone color determination model P23 acquires an image of the central part of the head (central image) from the training head images, and from this central image, it obtains five color space data: RGB, HSV, CIELAB (LAB), CIELUV (LUV), and XYZ. The image size of the central part of the head can be selected as appropriate by the machine learning model developer. A unified image size may be used, or it may be varied for each image.

[0145] In this embodiment, the Python® programming language library is used to acquire such color space data. Specifically, OpenCV, one of the Python libraries, is used to acquire numerical data of various color spaces from an image, and Numpy, another Python library, is used to convert the higher-order color space data (tensor) into lower-order data (matrix), as well as to perform statistical processing such as calculating the mean.

[0146] For example, suppose we obtain a 100x100 pixel image as the central part of the head from a training image.

[0147] RGB indicators have three parameters: R (Red), G (Green), and B (Blue). We will derive the values ​​for each of the RGB values ​​for the above 100 pixels x 100 pixels, totaling 10,000 pixels. For example, the R value will be derived for 10,000 pixels.

[0148] Next, we calculate the average value for each of R (Red), G (Green), and B (Blue). For example, we find the sum of 10,000 R values ​​and divide by 10,000. As a result, we obtain three numerical values, which are the average values ​​for each of RGB.

[0149] Similarly, parameters are obtained for HSV, CIELAB (LAB), CIELUV (LUV), and XYZ. Since each index has three parameters, 3 parameters × 5 indexes yield 15 features (color space data).

[0150] In other words, each image is associated with 15 features, along with a color category and tone determined by sensory evaluation, to form the training data.

[0151] In other words, the average value of each color space data is formatted as one-dimensional data to obtain 15 features corresponding to the correct labels for color and tone. In this embodiment, a total of 478 images are used as training data to train the hair tone color determination model P23.

[0152] Here, we will explain the reasons for using these five metrics. RGB is a widely used color space and is intuitively easy to use as a numerical value, so it is adopted. HSV uses Hue, Saturation, and Value as metrics, and has parameters such as saturation and value that RGB does not have, so it is adopted because it can distinguish colors that have the same hue but different brightness without correction, and it is expected to be able to take into account the effect of brightness. CIELAB (LAB) is adopted because it can express brightness etc. using the L axis (from black to white) as a metric. CIELUV (LUV) is adopted because it includes the concept of luminance and is expected to correct for vividness. XYZ is the parameter that CIELUV (LUV) and CIELAB (LAB) are based on, and it is adopted because a larger number of features makes it easier to grasp the characteristics of color. As a result of our evaluation, we adopted this combination of five because it provided good accuracy (color reproduction accuracy, etc.).

[0153] The bleaching condition provision program P24 is a program that outputs the bleaching conditions necessary to achieve the desired hair color after dyeing for those who wish to have their hair dyed.

[0154] In other words, the bleaching condition provision program P24 takes the hair damage level of the person requesting hair dyeing before bleaching, the tone before bleaching, and the target tone after dyeing as input, and outputs the bleaching conditions (name of the bleaching agent, concentration of the oxidizing agent during bleaching, processing time during bleaching, or number of bleaching sessions).

[0155] In this embodiment, the bleach condition provision program P24 takes the following as inputs: the hair damage level of the person requesting hair dyeing before bleaching, whether or not the natural hair is visible, the tone before bleaching, and the target tone after dyeing, and outputs the bleach conditions (name of the bleaching agent, root oxidizing agent concentration, which is the concentration of the oxidizing agent at the root of the hair when bleaching, tip oxidizing agent concentration, which is the concentration of the oxidizing agent at the tip of the hair when bleaching, root processing time, which is the processing time at the root of the hair when bleaching, tip processing time, which is the processing time at the tip of the hair when bleaching, or the number of bleaching cycles).

[0156] The bleach condition provision program P24 includes both a rule-based program for determining conditions and a program using a machine learning model. In the following, the rule-based program for determining conditions will be referred to as "rule-based program P24a," and the program using a machine learning model will be referred to as "machine learning program P24b." It is also called the bleach condition provision model P24b (bleach condition provision machine learning model P24b).

[0157] The rule-based program P24a will be described below using Figure 8 and Table 1. The machine learning program P24b will be described in the second embodiment described later.

[0158] Figure 8 shows the criteria for determining the bleaching conditions. The processor 122 obtains the necessary information (presence or absence of hair damage, target color tone after dyeing, and color tone before bleaching) through the bleaching condition provision process (Figure 7) up to step 6, and then obtains the bleaching conditions using the bleaching condition provision program P24. Details of the bleaching conditions will be explained later in the bleaching conditions table (Table 1).

[0159] First, the processor 122 determines whether or not there is hair damage (step 11). In this embodiment, the processor 122 obtains whether or not there is hair damage using the hair damage level acquisition process described above. In this embodiment, the description is based on whether or not there is hair damage, but for example, if the questionnaire regarding hair damage is a multi-level evaluation, the classification may be "hair damage is below a certain level (or above a certain level)".

[0160] After determining whether or not there is hair damage, the processor 122 determines whether or not the natural hair is visible (step 12 or step 13). In this embodiment, the processor 122 obtains whether or not the natural hair is visible by a natural hair visibility determination process using the natural hair visibility determination model P22 described above.

[0161] As shown in Figure 8, if there is hair damage (Step 11 Yes), proceed to the branch in Step 12; if there is no hair damage (Step 11 No), proceed to the branch in Step 13.

[0162] After determining whether or not the natural hair is visible, the processor 122 determines whether or not the target post-dye tone is equal to or greater than a predetermined numerical value (integer) (step 14, step 15, step 16, or step 17). In this embodiment, this predetermined numerical value is 12. However, this numerical value is the result of the investigation in this embodiment and may be changed as appropriate depending on the training data, etc. For convenience, this determination will be referred to as "determination of target post-dye tone". In this embodiment, the processor 122 obtains the target post-dye tone using the target post-dye tone acquisition process that uses the hair tone color determination model P23 described above.

[0163] As shown in Figure 8, if natural hair is visible, i.e., if step 12 Yes or step 13 Yes, the process proceeds to branching in step 14 or step 16, respectively. If natural hair is not visible, i.e., if step 12 No or step 13 No, the process proceeds to branching in step 15 or step 17, respectively.

[0164] Note that in each of the judgments in Figure 8, some may not require judgment and the process may proceed to the next branch. Examples will be explained in the next section on determining the tone before bleaching (steps 24 and 25).

[0165] After determining the target hair tone after dyeing, the processor 122 determines which of the predetermined categories the pre-bleach tone falls into (steps 18, 19, 20, 21, 22, 23, 24, or 25). For convenience, this determination is referred to as "pre-bleach tone determination".

[0166] Furthermore, as shown in step 24 or step 25 of Figure 8, depending on the branch, the bleaching conditions may be determined regardless of the pre-bleach tone value, but some parts of the pre-bleach tone determination may include cases where such determination is not required.

[0167] As shown in Figure 8, in determining the pre-bleach tone shown in step 18, the processor 122 outputs bleach condition 7 if the pre-bleach tone is 18 or higher, bleach condition 8 if the pre-bleach tone is 14 or higher but less than 18, and bleach condition 0 if the pre-bleach tone is less than 14. The bleach conditions (bleach condition 7, bleach condition 8, or bleach condition 0) will be described later (Table 1).

[0168] Similarly, in determining the pre-bleach tone as shown in step 19, the processor 122 outputs bleach condition 9 if the pre-bleach tone is 14 or higher, and bleach condition 6 if the pre-bleach tone is less than 14.

[0169] In determining the pre-bleach tone as shown in step 20, the processor 122 outputs bleach condition 5 if the pre-bleach tone is 18 or higher, and bleach condition 0 if the pre-bleach tone is less than 18.

[0170] In determining the pre-bleach tone as shown in step 21, the processor 122 outputs bleach condition 6 if the pre-bleach tone is 14 or higher, and bleach condition 0 if the pre-bleach tone is less than 14.

[0171] In determining the pre-bleach tone as shown in step 22, the processor 122 outputs bleach condition 2 if the pre-bleach tone is 18 or higher, and bleach condition 3 if the pre-bleach tone is less than 18.

[0172] In determining the pre-bleach tone as shown in step 23, the processor 122 outputs bleach condition 4 if the pre-bleach tone is 18 or higher, and bleach condition 0 if the pre-bleach tone is less than 18.

[0173] In determining the pre-bleach tone as shown in step 24, the processor 122 outputs bleach condition 0 regardless of the pre-bleach tone.

[0174] In determining the pre-bleach tone as shown in step 25, the processor 122 outputs bleach condition 1 regardless of the pre-bleach tone.

[0175] Here, I will provide some additional information regarding these branching points (oxidizing agent concentration and blending ratio).

[0176] First, let's explain the oxidizing agent concentration. For example, if the hair color category before bleaching is a light color (such as yellow) and there is hair damage (damage is above a certain level), the oxidizing agent concentration is set to 3.0%. Also, if the hair color category before bleaching is a light color (such as yellow) and there is no hair damage (damage is below a certain level), the oxidizing agent concentration is set to 4.5%. In all other cases, the oxidizing agent concentration is set to 6.0%.

[0177] Next, we will explain the mixing ratio of Part 1 (hair dye) and Part 2 (oxidizing agent) during hair dyeing. If the target hair tone after dyeing is 14 or higher, Part 2 (oxidizing agent) will be used in a ratio of three times that of Part 1 (hair dye). Similarly, if the target hair tone after dyeing is between 12 and 14, Part 2 (oxidizing agent) will be used in a ratio of twice that of Part 1 (hair dye), and if the target hair tone after dyeing is less than 12, the mixing ratio of Part 1 (hair dye) and Part 2 (oxidizing agent) will be the same (1:1).

[0178]

[0179] Table 1 is a bleaching conditions table. As shown in Table 1, the bleaching conditions table includes a unique bleaching condition number (bleaching condition ID), the number of bleaching sessions, and the bleaching conditions for the first and second bleaching sessions.

[0180] In this embodiment, the bleaching conditions include the following (1) to (4): (1) Root oxidizing agent concentration (referred to as "Root Ox concentration" in Table 1), which is the concentration of the oxidizing agent at the root of the hair during bleaching; (2) Root processing time (after application of the bleaching agent) during bleaching at the root of the hair; (3) Tip oxidizing agent concentration (referred to as "Tip Ox concentration" in Table 1), which is the concentration of the oxidizing agent at the tip of the hair during bleaching; (4) Tip processing time (after application of the bleaching agent) during bleaching at the tip of the hair.

[0181] The following explains the bleaching conditions for each type of bleach.

[0182] Bleach condition number 0 is for when there is no hair damage, no visible roots (not a so-called "root regrowth" state), and the target post-dye tone is 12 or higher.

[0183] Bleach condition number 1 is the bleaching condition when (1) there is no hair damage, no exposure of natural hair, and the target color tone after dyeing is less than 12, or (2) there is no hair damage, natural hair is exposed (so-called root growth), the pre-bleaching color tone is less than 14, and the target color tone after dyeing is less than 12.

[0184] Bleach condition number 2 is for cases where there is no hair damage, the natural hair is visible, the pre-bleach tone is 18 or higher, and the target post-dye tone is 12 or higher.

[0185] Bleaching condition number 3 is for cases where there is no hair damage, the natural hair is visible, the pre-bleach tone is 14 or higher but less than 18, and the target post-dye tone is 12 or higher.

[0186] Bleach condition number 4 is for cases where there is no hair damage, the natural hair is visible, the pre-bleach tone is 14 or higher but less than 18, and the target post-dye tone is less than 12.

[0187] Bleach condition number 5 is a bleaching condition for either (1) when there is hair damage, but the natural hair is not visible, and the target color tone after dyeing is 12 or higher, or (2) when there is hair damage, the natural hair is visible, and the target color tone after dyeing is 12 or higher.

[0188] Bleach condition number 6 is the bleach condition for (1) when there is hair damage, but the natural hair is not visible, and the target color tone after dyeing is less than 12, or (2) when there is hair damage, the natural hair is visible, and the target color tone after dyeing is less than 12.

[0189] Bleach condition number 7 is for cases where the hair is damaged, the natural hair color is visible, the pre-bleach tone is 18 or higher, and the target post-dye tone is 12 or higher.

[0190] Bleach condition number 8 is for cases where the hair is damaged, the natural hair color is visible, the pre-bleach tone is between 14 and 18, and the target post-dye tone is 12 or higher.

[0191] Bleach condition number 9 is for cases where the hair is damaged, the natural hair color is visible, the pre-bleach tone is between 14 and 18, and the target post-dye tone is less than 12.

[0192] Here, the bleach condition numbers in Figure 8 (from bleach condition 0 to bleach condition 9) correspond to the bleach condition numbers in Table 1. In other words, in this embodiment, the processor 122 obtains the bleach condition by obtaining the bleach condition number from the judgment conditions described above.

[0193] In this embodiment, the processor 122 prioritizes the conditions in Table 1, and if none of the conditions in Table 1 apply, it obtains the bleach condition with bleach condition number 0.

[0194] The user takes a photograph of the head (hair) of the person who wants their hair dyed and sends the bleached forehead image and hair damage level data to the server 10. The server 10 sends the bleaching conditions to the terminal 20, and the user refers to this information to bleach the hair of the person who wants their hair dyed.

[0195] Examples of bleaching agents include Bleach Master Cream manufactured by Shiseido Ltd., Charadeco Control Bleach manufactured by Chugai Pharmaceutical Co., Ltd., Charadeco Toner Bleach Powder EX manufactured by Chugai Pharmaceutical Co., Ltd., Ordeve Addicthy High Bleach manufactured by Milbon Corporation, Access Free Bleach manufactured by Napla Corporation, Blonder Bleach manufactured by Wella, and FIBREPLEX Bleach manufactured by Schwarzkopf Henkel. In addition, commercially available bleaching agents may be used.

[0196] <2-2. Hair Dye Condition Provisioning Process> The processor 122 performs the hair dye condition provisioning process based on the hair dye condition provisioning program P14. That is, the hair dye condition provisioning program P14 makes the computer function as a means of providing hair dye conditions by executing the hair dye condition provisioning process by the processor 122.

[0197] In the hair dyeing condition provision process, the processor 122 performs: a target post-dye color category acquisition process to acquire a target post-dye color category, which is the color category of the hair color after dyeing desired by the person requesting hair dyeing; a target post-dye tone acquisition process to acquire a target tone, which is a numerical value indicating the target brightness of the hair color after dyeing desired by the person requesting hair dyeing; a hair dyeing condition acquisition process to acquire hair dyeing conditions (including at least the type of hair dye and the waiting time during dyeing) from the target post-dye color category and the target post-dye tone; and a hair dyeing condition display process to display the hair dyeing conditions on the display unit 284a of the user's terminal (terminal 20).

[0198] Figure 9 is a flowchart showing the hair dyeing condition provision process. After the bleaching condition provision process, the processor 122 receives a request for hair dyeing conditions from the user and starts the hair dyeing condition provision process.

[0199] Processor 122 obtains the target color category after hair dyeing and the target tone after hair dyeing (step 31: process to obtain target color category after hair dyeing and process to obtain target tone after hair dyeing).

[0200] As described above, the processor 122 of this embodiment obtains the target post-dye color category by performing a target post-dye color category acquisition process when providing bleach conditions, and also obtains the target post-dye tone by performing a target post-dye tone acquisition process when providing bleach conditions.

[0201] Next, the processor 122 obtains the hair dyeing conditions (including at least the type of hair dye and the processing time during dyeing) from the acquired target post-dye color category and target post-dye tone (step 32: hair dyeing condition acquisition process).

[0202] The processor 122 then displays the acquired hair dyeing conditions on the display unit 284a of the user's terminal (terminal 20) (step 33: hair dyeing condition display process).

[0203] In this embodiment, the processor 122 executes the hair dyeing condition acquisition process using the hair dyeing condition provision program P25.

[0204] The hair dyeing condition provision program P25 includes a rule-based program and a machine learning model-based program. In the following, the rule-based program will be referred to as "rule-based program P25a," and the machine learning model-based program will be referred to as "machine learning program P25b."

[0205] The rule-based program P25a will be described below using Figure 8 and Table 1. The machine learning program P24b will be described in the third embodiment described later.

[0206] The rule-based program P25a uses a database in which target post-dye color categories and target post-dye tone correspond one-to-one with dyeing conditions. For example, if there are 12 color categories and 9 tones (levels) from 12 to 20, the database will have 12 × 9 = 108 dyeing conditions. Details will be explained in the database section (dyeing condition database D20). In other words, the processor 122 takes the target post-dye color category and target post-dye tone as input and outputs dyeing conditions using the rule-based program P25a.

[0207] The hair dyeing conditions include information on the type (name, brand name, etc.) of at least one type of hair dye, and information on the processing time during dyeing. If there are two or more types of hair dye, information on the amount (mixing ratio) of each hair dye is also included. In addition, the concentration of the oxidizing agent (oxidizer) used during dyeing and the mixing ratio of the oxidizing agent (oxidizer) to the hair dye are also included. Furthermore, if the dyeing conditions for the roots and ends of the hair differ, the dyeing conditions for both the roots and ends are also included.

[0208] After bleaching the hair of the person requesting hair dyeing, the user dyes (colors) the hair of the person requesting hair dyeing according to the hair dyeing conditions displayed in the hair dyeing conditions display process.

[0209] In this embodiment, the hair dyeing condition provision program P25 has the advantage of outputting hair dyeing conditions simply by inputting the target post-dye color category. This is based on the fact that the bleaching success rate is extremely high due to the bleaching condition provision program P24 described above, resulting in bleaching as planned.

[0210] <2-3. Hair Dyeing Technology Evaluation and Provision Processing> The processor 122 performs hair dyeing technology evaluation and provision processing based on the hair dyeing technology evaluation and provision program P16. That is, the hair dyeing technology evaluation and provision program P16 makes the computer function as a means of providing hair dyeing technology evaluation by the execution of the hair dyeing technology evaluation and provision processing by the processor 122.

[0211] In the hair dyeing technology evaluation provision process, the processor 122 executes: (a target post-dyeing color category acquisition process to acquire a target post-dyeing color category, which is the color category of the hair color after dyeing desired by the person requesting hair dyeing); a post-dyeing image acquisition process to acquire a post-dyeing image of the head, which is an image of the head of the person requesting hair dyeing after (actual) hair dyeing; an actual post-dyeing color category acquisition process to acquire an actual post-dyeing color category, which is the color category of the hair after dyeing, from the post-dyeing image of the post-dyeing head; an actual post-dyeing tone acquisition process to acquire an actual post-dyeing tone, which is a numerical value indicating the brightness of the hair after dyeing, from the post-dyeing image of the post-dyeing head; and a color reproduction evaluation process to output a color reproduction evaluation from the target post-dyeing color category, the target post-dyeing tone, the actual post-dyeing color category, and the actual post-dyeing tone; and a color reproduction display process to display the color reproduction evaluation on the display unit 284a of the terminal 20.

[0212] Figure 10 is a flowchart showing the hair dyeing technology evaluation and provision process. The processor 122 starts the hair dyeing technology evaluation and provision process when it receives a hair dyeing technology evaluation request and related data from the user. Receiving a hair dyeing technology evaluation request and related data means, for example, when the user sends an image of the back of their head with dyed hair via the terminal 20 and the server 10 receives it.

[0213] Processor 122 acquires a hair-dyeing image of the back of the head of a person who wishes to have their hair dyed, which is an image of the head after the actual hair dyeing (Step 41: Hair-dyeing back-of-head image acquisition process). This is so that Processor 122 can analyze the hair color of the person who wishes to have their hair dyed after the actual hair dyeing. As mentioned above, the hair color of the person who wishes to have their hair dyed after the actual hair dyeing is referred to as the "actual hair color after dyeing".

[0214] In this embodiment, the user takes a photograph of the head of a person who wishes to have their hair dyed after the dyeing process and sends it to the server 10. The processor 122 acquires this image information.

[0215] When detecting the image of the back of the head with dyed hair from the image taken by the imaging device of terminal 20, the head determination model P21 is used.

[0216] The processor 122 obtains the actual post-dye tone, which is a numerical value indicating the brightness of the hair after dyeing, from the image of the back of the head after dyeing (step 42: actual post-dye tone acquisition process).

[0217] The processor 122 also obtains the actual post-dye color category, which is the hair color category after dyeing, from the image of the back of the head after dyeing (step 42: actual post-dye color category acquisition process).

[0218] Specifically, the processor 122 uses the hair tone color determination model P23 to obtain the tone and color category of the hair after dyeing, targeting the central part of the head (center of the head) detected in the hair dyeing occipital head image acquired in step 41.

[0219] Although Figure 10 shows the actual post-dyeing tone and actual post-dyeing color category, the processor 122 acquires not only this data, but also other data necessary for evaluating the hair dyeing technique, which is covered in the next section on hair dyeing technique evaluation.

[0220] The processor 122 outputs an evaluation of the hair dyeing technique based on the various information it has acquired (step 43: hair dyeing technique evaluation process) and displays it on the display unit 284a of the terminal 20 (step 44: hair dyeing technique evaluation display process).

[0221] For example, if the hair dyeing technique evaluation is an evaluation of color reproduction accuracy, then the hair dyeing technique evaluation process will be the color reproduction accuracy evaluation process described later, and the hair dyeing technique evaluation display process will be the color reproduction accuracy display process.

[0222] Here, we will describe the evaluation of hair dyeing techniques in detail. As mentioned above, the high-tone color support system 1 of this embodiment evaluates hair dyeing techniques (procedures) through individual evaluations of color reproduction, gradation, and uniformity of application, as well as an overall evaluation called coloring evaluation.

[0223] "Color reproduction accuracy" is an index that evaluates, from the perspective of color category and tone, how well the actual hair color after dyeing (actual hair color after dyeing) of a person seeking hair dyeing is reproduced compared to the desired hair color after dyeing (image of hair color after dyeing / target hair color after dyeing). "Gradation" is an index that evaluates the vertical color gradation (uniformity of dyeing, smoothness of color change, or vividness of color change) of the actual hair color after dyeing (actual hair color after dyeing). "Uniformity of application" is an index that evaluates the horizontal uniformity of dyeing of the actual hair color after dyeing (actual hair color after dyeing).

[0224] In this embodiment, color reproduction accuracy, gradation, uniformity of coloring, and coloring evaluation are rated on a five-point scale. Although decimal numbers may appear in the calculation process, in this embodiment, integer values ​​are output for each evaluation by rounding. For example, an evaluation score from 2.5 to 3.4 is given a rating of 3. However, this is not limited to this, and evaluations in increments of 0.5 may also be used.

[0225] The following describes each evaluation method. In this embodiment, the evaluation (scoring) is performed using a point deduction system, but this is not the only method, and other methods such as a point addition system may be used for evaluation.

[0226] First, let's explain how to evaluate color reproduction accuracy. Color reproduction accuracy is determined from the color category and tone of the hair (before bleaching and after dyeing).

[0227] In other words, after the processor 122 retrieves the target post-dye color category, target post-dye tone, actual post-dye color category, and actual post-dye tone data stored in the memory unit 14, it evaluates the difference between the target post-dye color category and the actual post-dye color category, and the difference between the target post-dye tone and the actual post-dye tone. The processor 122 assigns a higher evaluation the smaller these differences are.

[0228] First, let's explain the evaluation of color categories. If the target hair color category after dyeing and the actual hair color category after dyeing match, 0 points will be deducted. If they are specific combinations that are close together on the color chart, such as "red" and "orange," 1 point will be deducted. If neither of these applies, 2 points will be deducted.

[0229] For example, in this embodiment, there is a color circle divided into 12 color categories. If the target post-dye color category and the actual post-dye color category are in the same color category, there is no penalty. If the target post-dye color category and the actual post-dye color category are separated by one, one point is deducted. If they are separated by two or more, two points are deducted.

[0230] However, this is not the only way to do so; more detailed classifications are also possible. Here, we will explain using two boundary values ​​M and N as examples.

[0231] The evaluation is performed using a color circle divided into multiple color categories. If the target post-dye color category and the actual post-dye color category are in the same color category, there is no penalty as described above. If the target post-dye color category and the actual post-dye color category are separated by 1 or more but less than M, 1 point is deducted. If they are separated by M or more but less than N, 2 points are deducted.

[0232] The values ​​of M and N here can be arbitrarily determined. Furthermore, there only needs to be one or more boundary values. For example, if there is one boundary value, it corresponds to the M mentioned above; if there are two boundary values, they correspond to both M and N mentioned above. There may also be three or more boundary values.

[0233] Furthermore, although this embodiment uses a point deduction system for evaluation, other methods such as a point addition system may also be used.

[0234] Next, we will explain the evaluation of tone. If the target tone after dyeing and the actual tone after dyeing match, 0 points are deducted. If the tones (levels) are a specific combination with close numerical values, such as "12" and "13", 1 point is deducted. If none of the above apply, 2 points are deducted.

[0235] While the evaluation of color categories uses a circular color chart, the evaluation of tone uses a number line; however, the evaluation methods are otherwise the same. In tone evaluation, the boundary values ​​are also the same as described above.

[0236] Next, we will explain how to evaluate gradients and the uniformity of fills.

[0237] Gradient refers to the color gradation of the hair in the vertical direction (upper part of the head), that is, from the roots (top of the head) to the tips. For example, a smooth gradient where the top of the head is slightly darker and the tips are lighter would be highly rated.

[0238] The uniformity of the color application refers to the uniformity of the hair color in the horizontal direction (lateral direction), for example, in areas roughly equidistant from the hair roots (top of the head). For example, a high rating is given when there are no uneven areas in the dye application.

[0239] Here, for evaluating the gradation and the uniformity of the coloring, the processor 122 uses the two-way hair dye evaluation program P26. The two-way hair dye evaluation program P26 comprises an evaluation area detection model P261, a color space data acquisition program P262, and a hair dye judgment model P263.

[0240] The evaluation region detection model P261 is a model that takes a head image (especially a back-of-head image of dyed hair) as input and outputs the evaluation region of the said head image.

[0241] The evaluation region detection model P261 first acquires an image of the back of a head with dyed hair and performs recognition of the head contained in that image.

[0242] The head recognition process is the same as that of the head determination (machine learning) model P21, but the processor 122 further acquires horizontal and vertical evaluation regions within the head region using the evaluation region detection model P261.

[0243] In this embodiment, the horizontal evaluation region is a horizontally elongated rectangular region, and the vertical evaluation region is a vertically elongated rectangular region.

[0244] If we assume that the head recognized by the evaluation region detection model P261 is egg-shaped, then the horizontal and vertical evaluation regions are the horizontally and vertically elongated rectangular regions contained within that egg-shaped region. Typically, the horizontal and vertical evaluation regions intersect to form a cross shape.

[0245] The evaluation region detection model P261 is a machine learning model that learns from images including the head (especially images of the back of a head with dyed hair), and the horizontal and vertical evaluation regions of the head in those images, and takes an image including the head (especially images of the back of a head with dyed hair) as input to output the horizontal and vertical evaluation regions of the head.

[0246] In other words, the processor 122 uses the evaluation region detection model P261 to take an image including the head (particularly an image of the back of the head with dyed hair) as input and executes an evaluation region detection process that outputs the horizontal evaluation region and the vertical evaluation region of the head.

[0247] Furthermore, the evaluation region detection model P261 has its training data pre-annotated to extract regions that do not contain background elements other than hair.

[0248] Furthermore, the evaluation region detection model P261 may be divided into a horizontal region detection model P261a that takes an image including the head (especially an image of the back of the head with dyed hair) as input and outputs (detects) only the horizontal evaluation region, and a vertical region detection model P261b that takes an image including the head (especially an image of the back of the head with dyed hair) as input and outputs (detects) only the vertical evaluation region.

[0249] Specifically, the processor 122 uses the horizontal region detection model P261a to perform a horizontal region detection process that takes an image including the head (especially an image of the back of the head with dyed hair) as input and outputs (detects) only the horizontal evaluation region. Similarly, the processor 122 uses the vertical region detection model P261b to perform a vertical region detection process that takes an image including the head (especially an image of the back of the head with dyed hair) as input and outputs (detects) only the vertical evaluation region.

[0250] Next, the processor 122 uses the color space data acquisition program P262 to acquire color space data in the horizontal and vertical evaluation regions and obtain statistical values.

[0251] The color space data acquisition program P262 takes a head image (especially a back-of-head image of dyed hair) as input and outputs the color space data for the head in that head image.

[0252] In other words, the processor 122, using the color space data acquisition program P262, takes a head image (especially an image of the back of the head with dyed hair) as input and executes a color space data acquisition process that outputs the color space data for the head in the head image.

[0253] The color space data acquisition program P262 of this embodiment is a program that takes a predetermined evaluation area (a rectangular area in this embodiment) of an image (particularly an image of the back of a head with dyed hair) as input and outputs color space data for said predetermined evaluation area.

[0254] In other words, the processor 122, using the color space data acquisition program P262, takes a predetermined evaluation area (a rectangular area in this embodiment) of an image (particularly an image of the back of a dyed head) as input and performs a color space data acquisition process that outputs color space data for the predetermined evaluation area.

[0255] Let's explain using the processing of the horizontal evaluation area as an example. Here, we assume that the horizontal evaluation area is 50 pixels high and 800 pixels wide.

[0256] First, processor 122 acquires color space data for all pixels (50 x 800 = 40,000 pixels).

[0257] In this embodiment, the color spaces are RGB, HSV, CIELAB (LAB), CIELUV (LUV), and XYZ. The color space data is data that quantifies each of these indices. In this embodiment, the color space data is extracted using OpenCV as described above.

[0258] Next, the processor 122 obtains the median and mean values ​​for each component and all components of the color space for each pixel in the vertical (column) direction.

[0259] For example, in the case of the RGB color space, each component of the color space is R (Red), G (Green), and B (Blue), and the total components are all of these components.

[0260] Let's take the example of a case where processor 122 obtains the average value of the R component of RGB for each pixel in the vertical direction (column direction).

[0261] Since we have obtained the R component value for all pixels in the vertical direction (50 pixels in this case), we calculate a single average value from these 50 R component values. In other words, we get one value for each column.

[0262] For simplicity, the value of the R component may be referred to as the "R value." The same applies to the component values ​​of each color space below. For abbreviations that are the same (for example, the L component (L value) of CIELAB (LAB) and CIELUV (LUV)), the meaning should be determined from the context.

[0263] Here, the horizontal evaluation area is 800 pixels (800 columns) wide, so the same process can be used to obtain the average value of 800 R components (arranged horizontally).

[0264] By performing the same process for the G and B components, we can obtain 800 average values ​​for the G components and 800 average values ​​for the B components.

[0265] In this embodiment, for RGB only, the average value of all components (RGB) is obtained from the average values ​​of the R component, the G component, and the B component. For example, if in a certain column the average value of the R component is 10, the average value of the G component is 40, and the average value of the B component is 40, then the average value of the RGB components is (10 + 40 + 40) ÷ 3 = 30. Since the average value of all components is also obtained for each column, in this example 800 RGB total component average values ​​are obtained.

[0266] The same applies to the median. In this example, we obtain the medians of 800 R components, 800 G components, 800 B components, and 800 RGB components.

[0267] Based on the above, mean and median data, each with 800 columns in a row, have been obtained for the R component, G component, B component, and all RGB components.

[0268] Next, the processor 122 obtains one of each of the following five parameters from the aforementioned 1 row and 800 columns of data: mean, median, standard deviation, coefficient of variation (standard deviation ÷ mean), and range (maximum value - minimum value). For example, the mean here is the sum of the 800 data points divided by 800.

[0269] To give a concrete example, from the 1 row, 800 columns of data relating to the mean of the R component mentioned above, we obtain one of each of the following five parameters: mean (referred to as the "overall mean" to distinguish it from the above), median (referred to as the "overall median" to distinguish it from the above), standard deviation, coefficient of variation (standard deviation ÷ mean), and range (maximum value - minimum value).

[0270] While RGB was used as an example here, in this embodiment, similar processing is applied to each component of the color spaces (indices) HSV, CIELAB (LAB), CIELUV (LUV), and XYZ in addition to RGB. For example, for HSV, the above processing is applied to the H (hue), S (saturation), and V (lightness) components. However, unlike RGB, the entire HSV spectrum is not used.

[0271] To generalize the processing for each indicator, if there is a horizontal evaluation area with a vertical dimension of a pixels and a horizontal dimension of y pixels, the mean and median values ​​of the color space data (including each component such as R (Red) and all components such as RGB) are obtained for each of the a pixels in the y columns, and a matrix data of 1 row and y columns is obtained. Then, the mean (overall mean), median (overall median), standard deviation, coefficient of variation, and range are obtained from the y data points in that 1 row and y column.

[0272] Furthermore, in addition to the above, processor 122 also acquires different features. This will be explained using the RGB example mentioned above (horizontal evaluation area of ​​50 pixels vertically and 800 pixels horizontally). Processor 122 identifies the color component with the largest numerical value and the color component with the smallest numerical value from the average value data of 800 columns per row (2,400 data points in total) for the R, G, and B components of RGB, and generates features using one-hot encoding.

[0273] Three variables are prepared for the color component with the largest numerical value. In this embodiment, these are "main_color_RGB_R", "main_color_RGB_G", and "main_color_RGB_B". Similarly, three variables are prepared for the color component with the smallest numerical value. In this embodiment, these are "minor_color_RGB_R", "minor_color_RGB_G", and "minor_color_RGB_B".

[0274] For example, if the average value of the R component is the largest, the variable main_color_RGB_R will be assigned 1, and main_color_RGB_G and main_color_RGB_B will be assigned 0.

[0275] On the other hand, if the average value of the G component is the largest, the variable minor_color_RGB_G is assigned a value of 1, and minor_color_RGB_R and minor_color_RGB_B are assigned a value of 0.

[0276] Similarly, if the average value of the R component is smallest, assign 1 to the variable minor_color_RGB_R, and 0 to minor_color_RGB_G and minor_color_RGB_B.

[0277] Here, one-hot encoding is applied only to the RGB and CIELAB (LAB) color spaces, for the color component with the largest and smallest numerical values. In other words, unlike other color spaces, processor 122 obtains six one-hot encoded features each for RGB and CIELAB (LAB).

[0278] As described above, the color space data acquisition program P262 outputs a total of 172 features (parameters) for the horizontal evaluation region from a single (post-dyeing) head image using the above process.

[0279] Specifically, for a total of 16 parameters—three components in RGB, all components of RGB, and three components each in HSV, CIELAB (LAB), CIELUV (LUV), and XYZ—the mean and median (2 parameters) are obtained for each, and furthermore, the overall mean, overall median, standard deviation, coefficient of variation, and range (5 parameters) are obtained. In other words, processor 122 obtains 16 × 2 × 5 = 160 parameters. Then, for RGB and CIELAB (LAB), six features each are obtained using the one-hot encoding method described above, for a total of 12 parameters, resulting in a total of 172 parameters.

[0280] The above explanation used a horizontal evaluation area as an example, but the same principle applies to the vertical evaluation area. A vertical evaluation area might be, for example, 800 pixels high and 50 pixels wide.

[0281] However, even if the hair is dyed correctly in the vertical direction, the roots (top of the head) will appear darker and the ends lighter in the photograph. Taking this into consideration, numerical processing such as fitting may be performed on the color space data obtained from the vertical evaluation region.

[0282] As described above, the color space data acquisition program P262 takes a predetermined evaluation region, that is, a head image (image of the back of the head with dyed hair) or at least a part of a head image (image of the back of the head with dyed hair) as input and outputs color space data of the head (evaluation region).

[0283] The color space data acquisition program P262 may be divided into a horizontal data acquisition program P262a and a vertical data acquisition program P262b.

[0284] In other words, the horizontal data acquisition program P262a takes a predetermined evaluation area, that is, a head image (image of the back of the head with dyed hair) or at least a part of a head image (especially horizontal data of the head) as input and outputs horizontal color space data of the head.

[0285] Similarly, the vertical data acquisition program P262b takes a predetermined evaluation area, i.e., a head image (image of the back of the head with dyed hair) or at least a portion of a head image (especially data in the vertical direction of the head) as input and outputs color space data in the vertical direction of the head.

[0286] The hair dye judgment model P263 is a machine learning model that learns using images of the back of a head with dyed hair and evaluations of the hair dyeing techniques related to those images as training data, and takes an image of the back of a head with dyed hair as input to output an evaluation of the hair dyeing techniques related to that image.

[0287] In other words, the processor 122 uses the hair dye determination model P263 to perform a hair dye determination process that takes a hair dyed back-of-head image as input and outputs an evaluation of the hair dyeing technique related to the said hair dyed back-of-head image.

[0288] Here, "evaluation of hair dyeing techniques related to images of the back of the head with dyed hair" refers to an evaluation (sensory evaluation) performed on the images of the back of the head with dyed hair by a beautician with advanced expertise in hair dyeing techniques. This evaluation includes an evaluation of the gradation (evaluation of hair dyeing techniques in the vertical direction of the hair) and an evaluation of the uniformity of the application (evaluation of hair dyeing techniques in the horizontal direction of the hair).

[0289] The hair dyeing judgment model P263 of this embodiment is a machine learning model that takes color space data in a predetermined region of a hair dyed occipital image as input and outputs an evaluation of the hair dyeing technique related to the hair dyed occipital image.

[0290] In this embodiment, the head image used for training the hair dye detection model P263 is the same as the head image used for training the evaluation region detection model P261.

[0291] The hair dyeing judgment model P263 includes a horizontal judgment model P263a and a vertical judgment model P263b. The horizontal judgment model P263a is a machine learning model that evaluates hair dyeing techniques in the horizontal direction of the hair (evaluation of uniformity of application), and the vertical judgment model P263b is a machine learning model that evaluates hair dyeing techniques in the vertical direction of the hair (evaluation of gradation).

[0292] In other words, the horizontal direction determination model P263a is a machine learning model that learns using a dyed hair occipital image (in this embodiment, color space data in a predetermined region (horizontal direction evaluation region) of the dyed hair occipital image obtained from the dyed hair occipital image) and an evaluation of the uniformity of the coloring (evaluation of the hair dyeing technique in the horizontal direction of the hair) as training data, and takes a dyed hair occipital image (in this embodiment, color space data in a predetermined region (horizontal direction evaluation region) of the dyed hair occipital image obtained from the dyed hair occipital image) as input and outputs an evaluation of the uniformity of the coloring (evaluation of the hair dyeing technique in the horizontal direction of the hair).

[0293] The processor 122 uses a horizontal direction determination model P263a to take a hair-dyed occipital image (in this embodiment, color space data in a predetermined region (horizontal direction evaluation region) of the hair-dyed occipital image obtained from the hair-dyed occipital image) as input and performs a horizontal direction determination process that outputs an evaluation of the uniformity of the coloring (evaluation of hair dyeing technique in the horizontal direction of the hair).

[0294] To give an example regarding the evaluation of color space data and uniformity of coloring in the horizontal evaluation domain, a small standard deviation in the color space data (e.g., the R value of RGB) in the horizontal evaluation domain tends to result in less uneven hair dyeing and a higher evaluation.

[0295] Furthermore, the vertical direction determination model P263b is a machine learning model that learns using a dyed hair occipital image (in this embodiment, color space data in a predetermined region (vertical direction evaluation region) of the dyed hair occipital image obtained from the dyed hair occipital image) and a gradient evaluation (evaluation of hair dyeing technique in the vertical direction of the hair) as training data, and takes a dyed hair occipital image (in this embodiment, color space data in a predetermined region (vertical direction evaluation region) of the dyed hair occipital image obtained from the dyed hair occipital image) as input and outputs a gradient evaluation (evaluation of hair dyeing technique in the vertical direction of the hair).

[0296] The processor 122 uses the vertical direction determination model P263b to take a hair-dyed occipital image (in this embodiment, color space data in a predetermined region (vertical direction evaluation region) of the hair-dyed occipital image obtained from the hair-dyed occipital image) as input and executes a vertical direction determination process that outputs a gradient evaluation (evaluation of hair dyeing technique in the vertical direction of the hair).

[0297] To give an example regarding the evaluation of color space data and gradients in the vertical evaluation domain, color space data in the vertical evaluation domain (for example, the R value of RGB) where the standard deviation, coefficient of variation, or range value falls within a certain range tends to result in a smooth gradient and a high evaluation. If these values ​​are too large, the gradient tends to be abrupt and color unevenness occurs. Conversely, if these values ​​are too small, the gradient is weak and appears flat.

[0298] In summary, the processor 122 performs the following processes: acquiring a hair-dyed occipital image, which is an image of the head of the person who wants their hair dyed after the hair dyeing process; a horizontal direction determination process (using the horizontal direction determination model P263a) that outputs an evaluation of the hair dyeing technique in the horizontal direction of the hair; and a vertical direction determination process (using the vertical direction determination model P263b) that outputs an evaluation of the hair dyeing technique in the vertical direction of the hair.

[0299] The horizontal direction determination model P263a is a machine learning model that learns color space data obtained from images of the back of the head with dyed hair and evaluations of hair dyeing techniques in the horizontal direction of the hair as training data, and outputs an evaluation of hair dyeing techniques in the horizontal direction of the hair when an image of the back of the head with dyed hair is input. The vertical direction determination model P263b is a machine learning model that learns color space data obtained from images of the back of the head with dyed hair and evaluations of hair dyeing techniques in the vertical direction of the hair as training data, and outputs an evaluation of hair dyeing techniques in the vertical direction of the hair when an image of the back of the head with dyed hair is input.

[0300] Finally, let's discuss the coloring evaluation. The coloring evaluation is the average of the evaluations for color reproduction accuracy, gradation, and uniformity of color application. In the example in Figure 6, the color reproduction accuracy is 2.0, the gradation is 3.0, and the uniformity of color application is 4.0, so the average of these values, 3.0, is the coloring evaluation.

[0301] With the above configuration, users performing bleaching and hair dyeing treatments on clients who desire high-tone hair color can learn the bleaching and dyeing conditions by using the High-Tone Color Support System 1. Since these bleaching and dyeing conditions are based on those set by hairdressers with advanced expertise in hair dyeing techniques, even users with limited experience in bleaching and hair dyeing treatments can perform appropriate bleaching and dyeing. Furthermore, by performing bleaching and dyeing using the bleaching and dyeing conditions provided by the High-Tone Color Support System 1, users can become proficient in these techniques.

[0302] As described above, the high-tone color assistance system 1 is equipped with various machine learning models. In particular, the two-way hair dyeing evaluation program P26 recognizes the head and determines the areas to be evaluated (horizontal evaluation area and vertical evaluation area), and acquires multiple color space data for those evaluation areas. Then, using this color space data as input, the evaluation model (horizontal judgment model and vertical judgment model) outputs an evaluation of the treatment. Users can obtain this evaluation for each treatment and improve their bleaching and hair dyeing techniques.

[0303] 3. Data Below, the data handled by the high-tone color support system 1 of this embodiment will be explained with reference to the figures. The high-tone color support system 1 of this embodiment is equipped with a high-tone color support database D1 in the storage unit 14 (data storage unit 14b) of the server 10. The high-tone color support database D1 is equipped with multiple databases, but here we will explain the bleach condition database D10, the hair dyeing condition database D20, and the treatment database D30.

[0304] The bleaching condition database D10 is a database that contains data related to bleaching conditions (bleaching condition data).

[0305]

[0306] Table 2 illustrates the data contained in the bleach condition database D10. Its contents correspond to those in Table 1 mentioned above.

[0307] As shown in Table 2, the bleach condition database D10 includes the following data items: a unique bleach condition ID (bleach condition number), the number of bleaching cycles, information on the bleaching agent (name and amount of bleaching agent), bleaching conditions (root oxidizing agent concentration, root processing time, tip oxidizing agent concentration, Emoto), and the name of the image data of the hair after bleaching is complete.

[0308] The post-bleaching image is the image that the processor 122 displays on the post-bleaching image display unit UI-32 of the bleaching condition display screen described above.

[0309] Although partially omitted in Table 2, if there are multiple bleaching sessions, the table includes data on the bleaching agent and bleaching conditions for each session.

[0310] The hair dyeing conditions database D20 is a database that contains data related to hair dyeing conditions (hair dyeing conditions data).

[0311]

[0312] Table 3 illustrates the data contained in the hair dyeing condition database D20. As shown in Table 3, the hair dyeing condition database D20 includes the following data items: a unique hair dyeing condition ID (hair dyeing condition number), a color category ID, a color category (after target hair dyeing), a tone (after target hair dyeing), information on the hair dye (name and amount of hair dye), hair dyeing conditions (processing time after application), the name of the image data for the completed hair dyeing image, and supplementary items.

[0313] The color category ID is a numerical ID that corresponds one-to-one with a color category. In this embodiment, the processor 122 can determine the hair dyeing conditions by obtaining the color category and tone.

[0314] The completed hair dyeing image is an image displayed by the processor 122 on the completed hair dyeing image display UI-42 of the hair dyeing condition display screen described above. Supplementary items include supplementary data related to color (light color, cool color, achromatic color, etc.).

[0315] Although omitted in Table 3, the hair dyeing conditions database D20 also contains further information regarding oxidizing agents (oxygen). For example, the concentration of the oxidizing agent (oxygen) and its mixing ratio with the dye.

[0316] The hair dyeing condition database D20 may include hair dyeing conditions other than those described above. For example, if the hair dyeing conditions for the roots and tips of the hair are different, the database may include both root and tip dyeing conditions. Also, if an oxidizing agent (oxygen) is required, the database may include the concentration of the oxidizing agent (oxygen concentration), etc.

[0317] The treatment database D30 is a database that contains data related to treatments (treatment data).

[0318]

[0319] Table 4 illustrates the data contained in the treatment database D30. As shown in Table 4, the hair dyeing condition database D30 includes the following data items: a unique treatment ID, a data name for the bleached forehead image, a data name for the post-dyeing image, a bleaching condition ID, a hair dyeing condition ID, a data name for the dyed back of the head image, and evaluation results (coloring evaluation, color reproduction, gradation, and uniformity of application).

[0320] The hair dyeing conditions database D30 may also contain customer data (such as customer IDs associated with customer personal information (name, address, contact information, visit history, number of times (beauty facility) has been used, etc.)). It may also contain information related to feedback (whether or not feedback was given and feedback comments) as described in the modified examples below.

[0321] For example, the processor 122 uses the treatment database D30 when displaying the hair dyeing technique evaluation screen or the analysis result display screen (not shown) as described above.

[0322] With the configuration described above, the High Tone Color Support Database D1 includes a database related to bleaching and hair dyeing conditions. Since this information can be updated, users can always access the latest bleaching and hair dyeing techniques.

[0323] 4. As shown in Hardware Configuration Diagram 1, the high-tone color assistance system 1 in this embodiment comprises a server 10 and a terminal 20. These devices are connected via a network N, which is, for example, the internet. The server 10 has software (application software) installed that includes the high-tone color assistance program P1 and is necessary to operate the high-tone color assistance system 1 according to this embodiment. Various processes are executed according to the functions of this software. The hardware components will be described below.

[0324] Figure 1 is a configuration diagram where server 10 is equipped with a high-tone color assistance program P1, and the high-tone color assistance system 1 is provided in the form of a web application. In contrast, there is also a case where terminal 20 is equipped with the high-tone color assistance program P1, and the high-tone color assistance system 1 is self-contained on terminal 20 and does not connect to network N. However, in this case, specific functions that can be obtained by connecting to network N, such as the use of updated machine learning models, cannot be used.

[0325] <Server 10> Server 10 is a computer for running the high-tone color support program P1. Although only one server 10 is shown in Figure 1, the number is not limited to one and may be implemented with multiple servers. For example, multiple servers may be used from the standpoint of load balancing and availability. Server 10 may utilize a computer provided by a cloud service provider, or the user may provide their own computer.

[0326] Figure 11 is a hardware configuration diagram of server 10. As shown in Figure 11, server 10 comprises a control unit 12, a storage unit 14, and a communication control unit 16. The control unit 12 also comprises a processor 122, ROM 124, RAM 126, and a timing unit 128. The basic functions of each will be explained in detail later.

[0327] The control unit 12, which includes the processor 122, also functions as a high-tone color auxiliary unit 130 in the server 10 (not shown). The high-tone color auxiliary unit 130 executes the high-tone color auxiliary program P1 to perform high-tone color auxiliary processing. In this embodiment, the processor 122 is a CPU (Central Processing Unit).

[0328] Furthermore, one program may include other programs. For example, in this embodiment, the high-tone color support program P1 includes a bleach condition provision program P12 and a hair dye condition provision program P14, etc.

[0329] As shown in Figure 11, the storage unit 14 includes a program storage unit 14a and a data storage unit 14b, and stores programs and data necessary for various processes. For example, the program storage unit 14a stores the high-tone color auxiliary program P1 according to this embodiment, as well as control programs for controlling devices connected to the server 10, such as a communication control program for controlling the communication control unit 16.

[0330] The communication control unit 16 is a device that communicates between the server 10 and an external terminal, such as terminal 20, which will be described later. As shown in Figure 1, the communication control unit 16 connects the server 10 to the network N.

[0331] In addition to the above, the server 10 may also be equipped with an input unit (e.g., a keyboard) for inputting commands and data, and an output unit (e.g., an audio output device) for outputting information in some form (not shown). Furthermore, it may be equipped with additional devices necessary for the use of this embodiment, or devices to improve convenience for the use of this embodiment.

[0332] <Terminal 20> Terminal 20 is an information processing device for users to use the high-tone color assistance system 1. Users can use the high-tone color assistance system 1 by accessing the server 10 using terminal 20.

[0333] Figure 12 is a hardware configuration diagram of terminal 20. As shown in Figure 12, terminal 20 comprises a control unit 22, a storage unit 24, a communication control unit 26, and an input / output unit 28. The control unit 22 also comprises a processor 222, ROM, RAM, and a timing unit. Although not shown in Figure 12, the input / output unit 28 comprises an input unit 282 and an output unit 284, and the output unit 284 comprises a display unit 284a. Explanations of content that overlaps with previously explained content or that relate to basic functions described later will be omitted.

[0334] In this embodiment, terminal 20 is a portable device such as a smartphone or tablet. However, terminal 20 is not limited to these, and may also be a stationary device such as a desktop PC.

[0335] The program storage unit 24a of the terminal 20 stores (installs) the user terminal application program P30 (user app P30) according to this embodiment, and the processor 222 executes various processes according to the functions of this software.

[0336] Various processes include output (screen display, audio output) based on information obtained from the server 10, receiving user input, and various communications. For example, when a user launches the user application P30, the processor 222 of the terminal 20 communicates with the server 10 via the network N and executes various processes.

[0337] In this embodiment, the user application P30 is installed on the terminal 20 via the network N or via a storage medium containing the user application P30.

[0338] (Explanation of the basic functions of the computer) The control unit (processor, ROM, RAM, timing unit), storage unit, communication control unit, input unit, and output unit will be described below. In any terminal of this embodiment, the connection configuration (network topology) between the functional units is not particularly limited. For example, it may be a bus type, star type, mesh type, etc.

[0339] The processor performs information processing and controls various devices according to programs stored in ROM or memory. In this embodiment, the processor is a CPU (Central Processing Unit).

[0340] However, the processor is not limited to a CPU. Various processors such as CPUs, DSPs (Digital Signal Units), GPUs (Graphics Processing Units), GPGPUs (General Purpose computing on GPUs), TPUs (Tensor Processing Units), or ASICs (Application Specific Integrated Circuits) may be used individually or in combination. For example, a processor that integrates a CPU and a GPU is called an APU (Accelerated Processing Unit), and such a processor may also be used.

[0341] ROM is read-only memory that contains various programs and data pre-stored for the processor to perform various control and calculations.

[0342] RAM is random-access memory used by the processor as working memory. Various areas can be reserved within this RAM for performing the various processes described in this embodiment.

[0343] The timing unit performs timing processing, including the acquisition of time information. If the computer is equipped with a communication control unit, it may acquire time information from an external source using NTP (Network Time Protocol).

[0344] A memory unit is a device for storing information such as programs and data. It is also called a storage unit. The memory unit can be either internal or external.

[0345] The storage unit includes a storage medium capable of reading and writing data, and a drive for reading and writing to the storage medium. Examples of storage media include internal and external types, such as HD (hard disk drive), CD-ROM, and flash memory. Examples of drives include HDD (hard disk drive) and SSD (solid state drive).

[0346] The memory unit comprises a program storage unit and a data storage unit as functional units. The program storage unit stores control programs for controlling various devices, such as communication control programs for controlling communications.

[0347] The communication control unit is a device for facilitating communication between terminals and other devices. The communication control unit connects the terminal equipped with the communication control unit to the network N.

[0348] The communication method used by the communication control unit is a known method, and either a wired or wireless method is applied depending on the device. For example, if the terminal is a desktop PC, both wired and wireless methods are possible, while if the terminal is a smartphone, a wireless communication method is possible.

[0349] For wired connections, communication methods specified in IEEE 802.3 (e.g., bus-type or star-type wired LANs) can be suitably used, but other communication methods such as those specified in IEEE 802.5 (e.g., ring-type wired LANs) may also be used.

[0350] For wireless communication, a communication method defined in IEEE 802.11 (e.g., Wi-Fi) can be suitably used. However, other methods such as IEEE 802.15 (e.g., Bluetooth®, BLE (Bluetooth® Low Energy)), IEEE 802.16 (e.g., WiMAX), ZigBee®, 920MHz band wireless (e.g., Wi-SUN), or optical communication methods such as infrared communication may also be used.

[0351] The input and output sections are devices responsible for input and output to the terminal, respectively. The input and output sections are sometimes collectively referred to as the input / output section. The input section is a device that receives input from the user. Examples of such input sections include keyboards, pointing devices such as mice, trackpads, tablets, or touch panels.

[0352] If the device is a tablet or smartphone and the input unit is a touch panel, the input unit is located on the surface of the display unit that displays images, such as a touchscreen. In this case, the input unit identifies the user's touch position corresponding to the various operation icons displayed on the display unit and accepts input from the user.

[0353] The output unit is a device for outputting images, audio, documents, etc. Examples of output units include display devices (display units) such as touchscreens and displays (liquid crystal displays and organic EL displays), audio output devices (audio output units) such as speakers, and document output devices (document output units) such as printers.

[0354] With the above configuration, users such as hairdressers can use the high-tone color support system 1 simply by installing the user application P30 on their own terminal (terminal 20).

[0355] (Second Embodiment) One of the bleach condition provision programs P24, the machine learning program P24b (bleach condition provision model P24b), is a machine learning model that outputs the bleach conditions necessary to achieve the desired hair color after dyeing for those who wish to dye their hair.

[0356] The machine learning program P24b learns the following data: hair damage level, tone before bleaching, target tone after dyeing, name of the bleaching agent, concentration of the oxidizing agent during bleaching, processing time during bleaching, and number of bleaching sessions.

[0357] As a result, the machine learning program P24b takes the hair damage level, pre-bleach tone, and target post-dye tone as input and outputs the bleaching conditions, namely the name of the bleaching agent, the concentration of the oxidizing agent during bleaching, the processing time during bleaching, and the number of bleaching cycles.

[0358] In this embodiment, the machine learning program P24b learns the following as training data: hair damage level, whether or not natural hair is visible, tone before bleaching, target tone after dyeing, root oxidizing agent concentration (the concentration of the oxidizing agent at the root of the hair during bleaching), tip oxidizing agent concentration (the concentration of the oxidizing agent at the tip of the hair during bleaching), root processing time (the processing time at the root of the hair during bleaching), tip processing time (the processing time at the tip of the hair during bleaching), and the number of bleaching sessions (at the root and tip of the hair, respectively).

[0359] As a result, the machine learning program P24b takes the hair damage level, whether or not the natural hair is visible, the tone before bleaching, and the target tone after dyeing as input and outputs the bleaching conditions, namely the oxidizing agent concentration at the roots, the oxidizing agent concentration at the ends, the processing time at the roots, the processing time at the ends, and the number of bleaching cycles (at the roots and ends of the hair, respectively).

[0360] In summary, the processor 122 performs the following: a hair damage level acquisition process to acquire the hair damage level of a person who wishes to dye their hair before bleaching; a bleached forehead image acquisition process to acquire a bleached forehead image, which is an image of the head of the person who wishes to dye their hair before bleaching; a bleached tone acquisition process to acquire a pre-bleach tone, which is a numerical value indicating the brightness of the hair before bleaching, from the bleached forehead image; a target post-dye tone acquisition process to acquire a target post-dye tone, which is a numerical value indicating the target brightness of the hair color after dyeing desired by the person who wishes to dye their hair; a bleach condition acquisition process to acquire bleach conditions including the name of the bleaching agent, the concentration of the oxidizing agent during bleaching, the waiting time during bleaching, and the number of bleaching sessions, using a bleach condition provision model (bleach condition provision model P24b); and a bleach condition display process to display the bleach conditions on the display unit of the user's terminal.

[0361] The bleach condition provision model (bleach condition provision model P24b) is a machine learning model characterized by learning the hair damage level before bleaching, the tone before bleaching, the target tone after dyeing, the name of the bleaching agent, the concentration of the oxidizing agent during bleaching, the processing time during bleaching, and the number of bleaching sessions as training data, and taking the hair damage level before bleaching, the tone before bleaching, and the target tone after dyeing as input, it outputs bleach conditions including the name of the bleaching agent, the concentration of the oxidizing agent during bleaching, the processing time during bleaching, and the number of bleaching sessions.

[0362] In other words, the bleach condition provision model (bleach condition provision model P24b) is a high-tone color assisting machine learning model characterized by learning the following as training data: the hair damage level of the person requesting hair dyeing before bleaching, the pre-bleach tone which is a numerical value indicating the brightness of the hair before bleaching obtained from the pre-bleach head image which is a head image of the person requesting hair dyeing before bleaching, the target post-dye tone which is a numerical value indicating the target brightness of the hair color after dyeing desired by the person requesting hair dyeing, the name of the bleaching agent, the oxidizing agent concentration during bleaching, the processing time during bleaching, and the number of bleaching sessions. By inputting the hair damage level of the person requesting hair dyeing before bleaching, the pre-bleach tone of the person requesting hair dyeing, and the target post-dye tone desired by the person requesting hair dyeing, it outputs the name of the bleaching agent, the oxidizing agent concentration during bleaching, the processing time during bleaching, and the number of bleaching sessions as bleaching conditions.

[0363] (Third Embodiment) One of the hair dyeing condition provision programs P25, the machine learning program P25b, is a machine learning model that outputs the hair dyeing conditions necessary to achieve the desired hair color after dyeing for the person who wants to dye their hair.

[0364] The machine learning program P25b learns the target post-dye hair color category, target post-dye hair tone, and dyeing conditions (such as the type of hair dye, the amount of dye used, or the processing time during dyeing) as training data.

[0365] Here, the hair dyeing conditions may include the amount (mixing ratio) of each hair dye used in the two or more hair dyes described above, as well as the hair dyeing conditions for the roots and ends of the hair.

[0366] As a result, the machine learning program P25b takes the target post-dye hair color category and target post-dye hair tone as input and outputs the hair dyeing conditions, such as the type of hair dye and the processing time during dyeing.

[0367] Here, the type of hair dye may be associated with color-related data (such as color categories). For example, the type of hair dye can be uniquely identified by the name of the hair dye (e.g., product name), and the color category corresponding to that hair dye is associated with it and stored in the memory of a computer equipped with a machine learning program P25b.

[0368] (Fourth Embodiment) The hair dyeing condition provision program P25 may also take as input data a post-bleach tone, which is a numerical value indicating the brightness of the hair of the person who wishes to dye their hair after bleaching, and / or a post-bleach color category, which is the color category of the hair of the person who wishes to dye their hair after bleaching.

[0369] In this case, the server 10 first obtains a photograph of the hair of a person who wishes to have their hair dyed after bleaching (bleached back of head image acquisition process), and then obtains the tone and color category of the hair (post-bleach tone acquisition process, post-bleach color category acquisition process).

[0370] For example, in the case of the machine learning program P25b, the following are learned as training data: post-bleach tone and / or post-bleach color category, target post-dye color category, and dyeing conditions (type of hair dye, amount of hair dye, and processing time during dyeing).

[0371] In this case, the hair dyeing condition provision program P25 takes the post-bleach tone and / or post-bleach color category and the target post-dye color category as input and outputs the hair dyeing conditions (type of hair dye, amount of hair dye, or processing time during dyeing).

[0372] In this embodiment, the hair tone color determination model P23 described above can be used to obtain the post-bleach tone and post-bleach color category.

[0373] By training the machine learning program P25b with post-bleaching data during its training phase, and by using the post-bleaching data as input during the inference phase, the high-tone color assistance system 1 can provide hair dyeing conditions with higher accuracy.

[0374] This is because the hair dyeing process involves bleaching followed by dyeing, making the information after bleaching more chronologically relevant to the information after dyeing than the information before bleaching.

[0375] (Modifications) The present invention is not limited to the embodiments described above, and includes various modifications to the embodiments described above, without departing from the spirit of the present invention.

[0376] For example, in the pre-bleach tone acquisition means (a pre-bleach tone acquisition means that acquires a pre-bleach tone, which is a numerical value indicating the brightness of the hair of a person who wishes to dye their hair before bleaching) and the target post-dye tone acquisition means (a target post-dye tone acquisition means that acquires a target tone, which is a numerical value indicating the target brightness of the hair color after dyeing desired by the person who wishes to dye their hair), the processor 122 may acquire a numerical value directly input by the user. In this case, the premise is that the high-tone color assistance system 1 is equipped with an interface for the user to directly input a numerical value.

[0377] The high-tone color support system 1 may include a feedback function for experts. This function involves sending data to experts via a network, including images of the bleached forehead and dyed back of the head, as well as at least some of the target post-dye color category, target post-dye tone, actual post-dye color category, actual post-dye tone, bleaching conditions, and dyeing conditions used for evaluating hair dyeing techniques, in a paid service or similar, and obtaining feedback from the experts who receive this information. In this embodiment, the feedback is in text format, but is not limited to text and may include images (including videos) or audio.

[0378] In other words, the server 10 may include an information sharing process that acquires information necessary for evaluating hair dyeing techniques, including at least a bleached forehead image and a dyed occipital head image, and transmits it to the expert's terminal, and a feedback transmission process that acquires feedback from the expert about the hair dyeing technique and transmits it to the user's terminal.

[0379] Furthermore, machine learning models may be used when evaluating the color reproduction accuracy as described above. Specifically, these may include: (1) a machine learning model for tone evaluation that learns target post-dye tone, actual post-dye tone, and color reproduction accuracy evaluation (related to tone) as training data, and outputs a color reproduction accuracy evaluation (related to tone) using target post-dye tone and actual post-dye tone as input; (2) a machine learning model for color category evaluation that learns target post-dye color category, actual post-dye color category, and color reproduction accuracy evaluation (related to color category) as training data, and outputs a color reproduction accuracy evaluation (related to color category) using target post-dye color category and actual post-dye color category as input; and (3) a machine learning model for color reproduction accuracy evaluation that learns target post-dye color category, target post-dye tone, actual post-dye color category, actual post-dye tone, and color reproduction accuracy evaluation as training data, and outputs a color reproduction accuracy evaluation using target post-dye color category, target post-dye tone, actual post-dye color category, and actual post-dye tone as input.

[0380] Aspects of the present invention, including this embodiment, have the following characteristics. The following corresponds to the claims at the time of filing the application. However, due to amendments to the claims after filing, the description of the claims after such amendments may differ. (1) In the first embodiment, a high-tone color assistance program is provided, characterized in that a computer functions as a hair damage level acquisition means for acquiring the hair damage level of a person who wishes to dye their hair before bleaching, a pre-bleach tone acquisition means for acquiring a pre-bleach tone which is a numerical value indicating the brightness of the hair of the person who wishes to dye their hair before bleaching, a target post-dye tone acquisition means for acquiring a target post-dye tone which is a numerical value indicating the target brightness of the post-dye hair color desired by the person who wishes to dye their hair, and a bleach condition acquisition means for acquiring bleach conditions including the name of the bleaching agent, the concentration of the oxidizing agent during bleaching, the waiting time during bleaching, or the number of bleaching sessions, taking the hair damage level, the pre-bleach tone, and the target post-dye tone as input. (2) In a second embodiment, a high-tone color assistance program is provided, characterized in that the computer functions as: a hair damage level acquisition means for acquiring the hair damage level of a person who wishes to dye their hair before bleaching; a bleached forehead image acquisition means for acquiring a bleached forehead image, which is a head image of the person who wishes to dye their hair before bleaching; a bleached tone acquisition means for acquiring a pre-bleach tone, which is a numerical value indicating the brightness of the hair before bleaching, from the bleached forehead image; a target post-dye tone acquisition means for acquiring a target post-dye tone, which is a numerical value indicating the target brightness of the hair color after dyeing desired by the person who wishes to dye their hair; a bleach condition acquisition means for acquiring bleach conditions, including the name of the bleaching agent, the concentration of the oxidizing agent during bleaching, the waiting time during bleaching, and the number of bleaching sessions, using the hair damage level, the pre-bleach tone, and the target post-dye tone as inputs; and a bleach condition display means for displaying the bleach conditions on the display unit of the user's terminal. In this case, the user can acquire more detailed bleach conditions, which has the advantage of improving the accuracy of the bleaching procedure.(3) In a third embodiment, the computer is further configured to function as a natural hair exposure determination means for determining whether or not natural hair is exposed from the bleached forehead image, and the bleach condition acquisition means is configured to take the hair damage level, whether or not natural hair is exposed, the tone before bleaching, and the target tone after dyeing as input and acquire bleach conditions including the name of the bleaching agent, the root oxidizing agent concentration which is the concentration of the oxidizing agent at the root of the hair during bleaching, the tip oxidizing agent concentration which is the concentration of the oxidizing agent at the tip of the hair during bleaching, the root processing time which is the processing time at the root of the hair during bleaching, the tip processing time which is the processing time at the tip of the hair during bleaching, and the number of bleaching cycles, thereby providing a high-tone color assistance program according to the second embodiment. In this case, the processor 122 provides detailed bleach conditions for both the root and tip of the hair, which has the advantage for the user of increased accuracy in the bleaching procedure. (4) In a fourth embodiment, the high-tone color assistance program described in the first embodiment is provided, characterized in that the computer functions as a target post-dye color category acquisition means for acquiring a target post-dye color category which is the color category of the hair color after dyeing desired by the person who wants to dye their hair; a hair dye condition acquisition means for acquiring hair dyeing conditions including at least the type of hair dye and the waiting time during dyeing from the target post-dye color category and the target post-dye tone; and a hair dye condition display means for displaying the hair dyeing conditions on the display unit of the user's terminal. In this case, the processor 122 provides the hair dyeing conditions, which has the advantage for the user of increased accuracy in the hair dyeing procedure.(5) In a fifth embodiment, the high-tone color assistance program described in the first embodiment is provided, characterized in that the computer functions as a target post-dye color category acquisition means for acquiring a target post-dye color category, which is the color category of the hair color after dyeing desired by the person requesting to dye their hair; a post-dye image acquisition means for acquiring a post-dye image of the head, which is a head image of the person requesting to dye their hair after dyeing; an actual post-dye color category acquisition means for acquiring an actual post-dye color category, which is the color category of the hair after dyeing, from the post-dye image of the head; an actual post-dye tone acquisition means for acquiring an actual post-dye tone, which is a numerical value indicating the brightness of the hair after dyeing, from the post-dye image of the head; and a color reproduction evaluation means for outputting an evaluation of color reproduction from the target post-dye color category, the target post-dye tone, the actual post-dye color category, and the actual post-dye tone. In this case, the processor 122 provides an evaluation of the color reproduction after dyeing, which has the advantage for the user of being able to improve their hair dyeing technique (especially the adjustment of the color tone). (6) In the sixth embodiment, the computer is further configured to function as a hair-dyeing occipital head image acquisition means for acquiring a hair-dyeing occipital head image, which is an image of the head of the person who wishes to have their hair dyed after the hair dyeing; a horizontal direction evaluation means for outputting an evaluation of the hair dyeing technique in the horizontal direction of the hair using a horizontal direction determination model; and a vertical direction evaluation means for outputting an evaluation of the hair dyeing technique in the vertical direction of the hair using a vertical direction determination model, wherein the horizontal direction determination model is a machine learning model that learns color space data obtained from the hair-dyeing occipital head image and an evaluation of the hair dyeing technique in the horizontal direction of the hair as training data, and outputs an evaluation of the hair dyeing technique in the horizontal direction of the hair using the hair-dyeing occipital head image as input; and the vertical direction determination model is a machine learning model that learns color space data obtained from the hair-dyeing occipital head image and an evaluation of the hair dyeing technique in the vertical direction of the hair as training data, and outputs an evaluation of the hair dyeing technique in the vertical direction of the hair using the hair-dyeing occipital head image as input, thereby providing the high-tone color assistance program described in the first embodiment. In this case, the processor 122 provides evaluations of hair dyeing techniques in the horizontal and vertical directions, which has the advantage for the user of being able to improve hair dyeing techniques (especially the uniformity of the application of hair dye).(7) In the seventh aspect, the computer functions as a hair damage level acquisition means for acquiring the hair damage level of a person who wishes to dye their hair before bleaching, a bleached forehead image acquisition means for acquiring a bleached forehead image which is a head image of the person who wishes to dye their hair before bleaching, a bleached forehead image acquisition means for acquiring a pre-bleach tone which is a numerical value indicating the brightness of the hair before bleaching from the bleached forehead image, a target post-dye tone acquisition means for acquiring a target post-dye tone which is a numerical value indicating the target brightness of the hair color after dyeing desired by the person who wishes to dye their hair, a bleach condition acquisition means for acquiring bleach conditions including the name of the bleaching agent, the concentration of the oxidizing agent during bleaching, the waiting time during bleaching, and the number of bleaching sessions, using a bleach condition provision model, and a bleach condition display means for displaying the bleach conditions on the display unit of the user's terminal, wherein the bleach condition provision model learns the hair damage level before bleaching, the pre-bleach tone, the target post-dye tone, the name of the bleaching agent, the concentration of the oxidizing agent during bleaching, the waiting time during bleaching, and the number of bleaching sessions as training data, This invention provides a high-tone color assistance program characterized by being a machine learning model that takes the hair damage level before bleaching, the tone before bleaching, and the target tone after dyeing as input, and outputs bleaching conditions including the name of the bleaching agent, the concentration of the oxidizing agent during bleaching, the processing time during bleaching, and the number of bleaching cycles. (8) In the eighth aspect, a high-tone color assisting machine learning model is provided, characterized in that the model learns the following as training data: the hair damage level of the person requesting hair dyeing before bleaching, the pre-bleach tone which is a numerical value indicating the brightness of the hair before bleaching obtained from a pre-bleach image which is a head image of the person requesting hair dyeing before bleaching, the target post-dye tone which is a numerical value indicating the target brightness of the hair color after dyeing desired by the person requesting hair dyeing, the name of the bleaching agent, the concentration of the oxidizing agent during bleaching, the processing time during bleaching, and the number of bleaching sessions. By inputting the hair damage level of the person requesting hair dyeing before bleaching, the pre-bleach tone of the person requesting hair dyeing, and the target post-dye tone desired by the person requesting hair dyeing, the model outputs the name of the bleaching agent, the concentration of the oxidizing agent during bleaching, the processing time during bleaching, and the number of bleaching sessions as bleaching conditions.(9) The ninth embodiment provides a high-tone color assistance system comprising: a hair damage level acquisition unit that acquires the hair damage level of a person who wishes to have their hair dyed before bleaching; a pre-bleach tone acquisition unit that acquires a pre-bleach tone which is a numerical value indicating the brightness of the person's hair before bleaching; a target post-bleach tone acquisition unit that acquires a target post-bleach tone which is a numerical value indicating the target brightness of the hair color desired by the person who wishes to have their hair dyed; and a bleach condition acquisition unit that takes the hair damage level, the pre-bleach tone, and the target post-bleach tone as inputs and acquires bleach conditions including any of the following: the name of the bleaching agent, the concentration of the oxidizing agent during bleaching, the waiting time during bleaching, or the number of bleaching sessions. (10) In the tenth aspect, a high-tone color assistance method is provided, characterized in that the processor comprises: a hair damage level acquisition step of acquiring the hair damage level of a person who wishes to have their hair dyed before bleaching; a pre-bleach tone acquisition step of the processor acquiring a pre-bleach tone which is a numerical value indicating the brightness of the hair of the person who wishes to have their hair dyed before bleaching; a target post-bleach tone acquisition step of the processor acquiring a target post-bleach tone which is a numerical value indicating the target brightness of the hair color desired by the person who wishes to have their hair dyed; and a bleach condition acquisition step of the processor taking the hair damage level, the pre-bleach tone, and the target post-bleach tone as input to acquire bleach conditions including the name of the bleaching agent, the concentration of the oxidizing agent during bleaching, the waiting time during bleaching, or the number of bleaching sessions.

[0381] As data accumulates, machine learning models will improve, allowing for more precise responses to customer needs (those wanting their hair dyed). Additionally, an increase in salons and other businesses capable of handling high-tone colors will broaden people's fashion choices, leading to further development of the beauty industry.

[0382] 1: High-tone color assist system 10: Server (12: Control unit (122: Processor, 124: ROM, 126: RAM, 128: Timing unit, 130: High-tone color assist unit), 14: Storage unit (14a: Program storage unit, 14b: Data storage unit), 16: Communication control unit 18: Input / output unit) 20: Terminal (22: Control unit, 24: Storage unit, 26: Communication control unit, 28: Input / output unit (282: Input unit, 284: Output unit (284a: Display unit))) UI-11: Real-time image display unit (UI-111: Head guide (UI-111a: Left / right guide, UI-111b: Front / back guide)), UI-12: Shooting button UI-21: Post-dye image display unit, UI-22: Operation UI display unit UI-31: Bleach condition display unit, UI-32: Image display unit after bleaching is complete UI-41: Hair dyeing condition display unit, UI-42: Hair dyeing completion image display unit UI-51: Evaluation display unit, UI-52: Analysis result link unit P1: High-tone color support program P12: Bleach condition provision program, P14: Hair dyeing condition provision program, P16: Hair dyeing technique evaluation provision program P21: Head judgment (machine learning) model P22: Natural hair exposure judgment (machine learning) model P23: Hair tone color judgment (machine learning) model P24: Bleach condition provision program (P24a: Rule-based program, P24b: Machine learning program (bleach condition provision model)) P25: Hair dyeing condition provision program (P25a: Rule-based program, P25b: Machine learning program) P26: Two-way hair dyeing evaluation program P261: Evaluation area detection (machine learning) model P261a: Horizontal area detection model P261b: Vertical area detection model P262: Color space data acquisition program P262a: Horizontal data acquisition program P262b: Vertical data acquisition program P263: Hair dye judgment (machine learning) model P263a: Horizontal judgment (machine learning) model P263b: Vertical judgment (machine learning) model P30: User terminal application program (user app) D1: High-tone color support database D10: Bleach condition database, D20: Hair dye condition database, D30: Treatment database

Claims

1. A high-tone color assistance program characterized by causing a computer to function as: a hair damage level acquisition means for acquiring the hair damage level of a person who wishes to have their hair dyed before bleaching; a pre-bleach tone acquisition means for acquiring a pre-bleach tone, which is a numerical value indicating the brightness of the person's hair before bleaching; a target post-bleach tone acquisition means for acquiring a target post-bleach tone, which is a numerical value indicating the target brightness of the desired post-bleach hair color; and a bleach condition acquisition means for acquiring bleach conditions, which include the name of the bleaching agent, the concentration of the oxidizing agent during bleaching, the waiting time during bleaching, or the number of bleaching sessions, using the hair damage level, the pre-bleach tone, and the target post-bleach tone as input.

2. A high-tone color assistance program characterized in that it causes a computer to function as: a hair damage level acquisition means for acquiring the hair damage level of a person who wishes to dye their hair before bleaching; a bleached forehead image acquisition means for acquiring a bleached forehead image, which is an image of the head of the person who wishes to dye their hair before bleaching; a bleached tone acquisition means for acquiring a pre-bleach tone, which is a numerical value indicating the brightness of the hair before bleaching, from the bleached forehead image; a target post-dye tone acquisition means for acquiring a target post-dye tone, which is a numerical value indicating the target brightness of the hair color after dyeing desired by the person who wishes to dye their hair; a bleach condition acquisition means for acquiring bleach conditions, including the name of the bleaching agent, the concentration of the oxidizing agent during bleaching, the waiting time during bleaching, and the number of bleaching sessions, using the hair damage level, the pre-bleach tone, and the target post-dye tone as input; and a bleach condition display means for displaying the bleach conditions on the display unit of the user's terminal.

3. The high-tone color assistance program according to claim 2, further comprising: a computer functioning as a natural hair exposure determination means for determining whether or not natural hair is exposed from the bleached forehead image; and a bleach condition acquisition means that takes the hair damage level, whether or not natural hair is exposed, the tone before bleaching, and the target tone after dyeing as input and acquires bleach conditions including the name of the bleaching agent, the root oxidizing agent concentration which is the concentration of the oxidizing agent at the root of the hair during bleaching, the tip oxidizing agent concentration which is the concentration of the oxidizing agent at the tip of the hair during bleaching, the root processing time which is the processing time at the root of the hair during bleaching, the tip processing time which is the processing time at the tip of the hair during bleaching.

4. The high-tone color assistance program according to claim 1, further characterized in that the computer functions as: a target post-dye color category acquisition means for acquiring a target post-dye color category which is the color category of the hair color after dyeing desired by the person who wants to dye their hair; a hair dye condition acquisition means for acquiring hair dyeing conditions including at least the type of hair dye and the waiting time during dyeing from the target post-dye color category and the target post-dye tone; and a hair dye condition display means for displaying the hair dyeing conditions on the display unit of the user's terminal.

5. The high-tone color assistance program according to claim 1, further characterized in that the computer functions as: a target post-dye color category acquisition means for acquiring a target post-dye color category, which is the color category of the hair color after dyeing desired by the person requesting hair dyeing; a post-dye image acquisition means for acquiring a post-dye image, which is a head image of the person requesting hair dyeing after dyeing; an actual post-dye color category acquisition means for acquiring an actual post-dye color category, which is the color category of the hair after dyeing, from the post-dye image; an actual post-dye tone acquisition means for acquiring an actual post-dye tone, which is a numerical value indicating the brightness of the hair after dyeing, from the post-dye image; and a color reproduction evaluation means for outputting a color reproduction evaluation from the target post-dye color category, the target post-dye tone, the actual post-dye color category, and the actual post-dye tone.

6. Furthermore, the computer functions as: a hair-dyeing occipital image acquisition means for acquiring a hair-dyeing occipital image, which is an image of the head of the person who wishes to have their hair dyed after the hair dyeing process; a horizontal direction evaluation means for outputting an evaluation of the hair dyeing technique in the horizontal direction of the hair using a horizontal direction determination model; and a vertical direction evaluation means for outputting an evaluation of the hair dyeing technique in the vertical direction of the hair using a vertical direction determination model; wherein the horizontal direction determination model is a machine learning model that learns color space data obtained from the hair-dyeing occipital image and an evaluation of the hair dyeing technique in the horizontal direction of the hair as training data, and outputs an evaluation of the hair dyeing technique in the horizontal direction of the hair using the hair-dyeing occipital image as input; and the vertical direction determination model is a machine learning model that learns color space data obtained from the hair-dyeing occipital image and an evaluation of the hair dyeing technique in the vertical direction of the hair as training data, and outputs an evaluation of the hair dyeing technique in the vertical direction of the hair using the hair-dyeing occipital image as input; the high-tone color assistance program according to claim 1.

7. The computer is configured to function as: a hair damage level acquisition means for acquiring the hair damage level of a person requesting hair dyeing before bleaching; a bleached forehead image acquisition means for acquiring a bleached forehead image, which is a head image of the person requesting hair dyeing before bleaching; a bleached tone acquisition means for acquiring a pre-bleach tone, which is a numerical value indicating the brightness of the hair before bleaching, from the bleached forehead image; a target post-dye tone acquisition means for acquiring a target post-dye tone, which is a numerical value indicating the target brightness of the hair color after dyeing desired by the person requesting hair dyeing; a bleach condition acquisition means for acquiring bleach conditions including the name of the bleaching agent, the concentration of the oxidizing agent during bleaching, the waiting time during bleaching, and the number of bleaching sessions, using a bleach condition provision model; and a bleach condition display means for displaying the bleach conditions on the display unit of the user's terminal, wherein the bleach condition provision model learns the hair damage level before bleaching, the pre-bleach tone, the target post-dye tone, the name of the bleaching agent, the concentration of the oxidizing agent during bleaching, the waiting time during bleaching, and the number of bleaching sessions as training data. A high-tone color assistance program characterized by being a machine learning model that takes the hair damage level before bleaching, the tone before bleaching, and the target tone after dyeing as input, and outputs bleaching conditions including the name of the bleaching agent, the concentration of the oxidizing agent during bleaching, the processing time during bleaching, and the number of bleaching sessions.

8. A machine learning model for assisting high-tone coloring, characterized in that it learns the following as training data: the hair damage level of a person who wishes to dye their hair before bleaching; the pre-bleach tone, which is a numerical value indicating the brightness of the hair before bleaching obtained from a pre-bleach head image of the person who wishes to dye their hair before bleaching; the target post-dye tone, which is a numerical value indicating the target brightness of the hair color after dyeing desired by the person who wishes to dye their hair; the name of the bleaching agent; the concentration of the oxidizing agent during bleaching; the processing time during bleaching; and the number of bleaching sessions; and outputs the name of the bleaching agent, the concentration of the oxidizing agent during bleaching; the processing time during bleaching; and the number of bleaching sessions as bleaching conditions when the hair damage level of a person who wishes to dye their hair before bleaching; the pre-bleach tone of a person who wishes to dye their hair; and the target post-dye tone desired by the person who wishes to dye their hair.

9. A high-tone color assistance system comprising: a hair damage level acquisition unit for acquiring the hair damage level of a person who wishes to have their hair dyed before bleaching; a pre-bleach tone acquisition unit for acquiring a pre-bleach tone, which is a numerical value indicating the brightness of the person's hair before bleaching; a target post-bleach tone acquisition unit for acquiring a target post-bleach tone, which is a numerical value indicating the target brightness of the hair color desired by the person who wishes to have their hair dyed; and a bleach condition acquisition unit that takes the hair damage level, the pre-bleach tone, and the target post-bleach tone as input and acquires bleach conditions including one of the following: the name of the bleaching agent, the concentration of the oxidizing agent during bleaching, the waiting time during bleaching, or the number of bleaching sessions.

10. A method for assisting with high-tone hair coloring, characterized in that the processor comprises: a hair damage level acquisition step in which the processor acquires the hair damage level of a person who wishes to have their hair dyed before bleaching; a pre-bleach tone acquisition step in which the processor acquires a pre-bleach tone, which is a numerical value indicating the brightness of the hair of the person who wishes to have their hair dyed before bleaching; a target post-bleach tone acquisition step in which the processor acquires a target post-bleach tone, which is a numerical value indicating the target brightness of the hair color desired by the person who wishes to have their hair dyed; and a bleach condition acquisition step in which the processor takes the hair damage level, the pre-bleach tone, and the target post-bleach tone as input and acquires bleach conditions including one of the following: the name of the bleaching agent, the concentration of the oxidizing agent during bleaching, the waiting time during bleaching, or the number of bleaching sessions.

Images