JP2018202044A - Method for evaluating clogging of pores and device for evaluating clogging of pores - Google Patents

Abstract

To provide a technique that can accurately evaluate clogging of pores.SOLUTION: A method for evaluating clogging of pores has an acquisition step (S13) that acquires a blue component image from a skin image captured in irradiation of light including at least a part of a blue wavelength range, and an evaluation step (S20) that evaluates a state of keratotic plugs on the basis of the acquired blue component image.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a skin image analysis technique, and more particularly to a pore clogging evaluation technique using a skin image.

Japanese Patent Application Laid-Open No. 2004-133826 shoots human skin by irradiating each of first and second light beams having different wavelength ranges, detects pores from the obtained first skin image, An apparatus is disclosed that detects porphyrins and porphyrin precursors from the skin image and displays an image showing the state of pores in human skin by distinguishing between pores in which an abnormality has been detected and normal pores.
Patent Document 2 discloses a skin condition analysis method in which pore analysis is performed using an image photographed by flash, and porphyrin analysis of a horn plug is performed using an image photographed by receiving visible light under ultraviolet irradiation. . In the pore analysis, the G component or R component image of the RGB component is used to extract and analyze a small shadowed portion as a pore. In the porphyrin analysis of the square plug, the number and the total area of the fluorescence considered to be derived from porphyrin are quantified using the R component image of the RGB component.
In Patent Document 3, a face is photographed in a state of being irradiated with ultraviolet rays, a specific color component (described as being particularly preferable for the R component) is extracted from the obtained face image, and the extracted color component is determined in advance. A method for detecting porphyrin by performing binarization with a threshold value of is disclosed.

International Publication No. 2015/159732 JP 2007-152084 A JP 2009-494 A

The methods disclosed in Patent Documents 1, 2, and 3 described above detect the porphyrin fluorescent region present in the pores from the skin image captured under ultraviolet irradiation. According to Patent Documents 1 and 2, pores in which porphyrin is fluorescent can be distinguished from pores that are not.
Porphyrin is a metabolite of acne bacteria that prefers sebum, and since keratin plugs contain sebum, porphyrins should basically be present in pores clogged with horn plugs.
However, there are pores that are blocked by horn plugs and in which porphyrin fluorescence is not properly imaged, and pores in which porphyrin remains even though the horn plugs are removed. That is, the clogged state of pores and the fluorescence state of porphyrin do not completely coincide with each other, and clogging of pores cannot be accurately evaluated by the above-described method.
The present invention relates to a technique capable of evaluating clogging of pores with high accuracy.

In the aspect of the present invention, the following configurations are employed in order to solve the above-described problems.
An aspect of the present invention relates to a pore clogging evaluation method. A pore clogging evaluation method includes an acquisition step of acquiring a blue component image from a skin image captured under irradiation of light including at least a part of a blue wavelength region, and a state of a horn plug based on the acquired blue component image And an evaluation step for evaluating.
Another aspect of the present invention relates to, for example, a pore clogging evaluation apparatus (information processing apparatus, computer) that executes the pore clogging evaluation method according to the above aspect, and the computer performs the pore clogging evaluation method according to the above aspect. In particular, the present invention relates to a computer-readable storage medium storing such a program. This recording medium includes a non-transitory tangible medium.

  According to the said aspect, the technique which can evaluate clogging of a pore with high precision can be provided.

It is a flowchart which shows the pore clogging evaluation method which concerns on this embodiment. It is a figure which shows notionally the hardware structural example of the pore clogging evaluation apparatus (evaluation apparatus) which concerns on this embodiment. It is a figure which shows notionally the process structural example of the pore clogging evaluation apparatus (evaluation apparatus) which concerns on this embodiment. FIG. 4A is a diagram illustrating an example of an original skin image, and FIG. 4B is a diagram illustrating an example of a ground truth skin image. It is a graph which shows the relationship between the area of R square plug area | region, and the long diameter of a square plug. It is a graph which shows the relationship between the area of a G square plug area | region, and the long diameter of a square plug. It is a graph which shows the relationship between the area of a B square plug area | region, and the long diameter of a square plug. It is a graph which shows the relationship between the area of the RGB square plug area | region as a comparative example, and the long diameter of a square plug.

  Hereinafter, an example of a preferred embodiment of the present invention (hereinafter referred to as this embodiment) will be described. In addition, embodiment mentioned below is an illustration and this invention is not limited to the structure of the following embodiment.

First, the outline | summary of the clogging evaluation method (henceforth this method) which concerns on this embodiment is demonstrated.
This method roughly includes an acquisition step of acquiring a blue component image from the skin image, and an evaluation step of evaluating the state of the horn plug based on the blue component image acquired in the acquisition step.

A “skin image” is under irradiation of light including at least a part of the blue wavelength region, specifically, light including a spectrum peak (spectral component) whose power at 390 nm is 10% or more of the maximum value. It is the image by which the test subject's skin was imaged. In the skin image, a portion of the subject's skin where it is desired to evaluate pore clogging is copied. For example, the skin image can be obtained by capturing the skin with a digital camera in an environment in which human skin is irradiated with light including at least a part of the blue wavelength region. As the digital camera at this time, a general digital camera that captures visible light is used.
The “blue wavelength region” is a wavelength band extending from the boundary between the ultraviolet wavelength band and the visible light band to the wavelength band of blue component light called blue light. In the present embodiment, the wavelength serving as the boundary is not limited. For example, Japanese Industrial Standard JIS Z 8120 stipulates that the short wavelength limit of the visible light wavelength range may be 360 to 400 nm, and the long wavelength limit may be 760 to 830 nm. For this reason, the “blue wavelength region” is set to, for example, 380 nm to 500 nm.
The light irradiated in the imaging environment of the skin image may be light that includes at least a part of such a blue wavelength region, and is a single color light (blue light) having only the blue wavelength region as a spectral component. It may be light mainly composed of part of the blue wavelength region and part of the green wavelength region longer than the blue wavelength region, or part of the blue wavelength region and the ultraviolet wavelength. Although the light may be mainly composed of a part of the region, the peak of the spectrum (spectrum component) in which the power at, for example, 390 nm at the boundary between the ultraviolet wavelength band and the visible light band is 10% or more of the maximum value. It is preferable that the light is mainly composed of a part of the blue wavelength region including the light and a part of the ultraviolet wavelength region.

  The “blue component image” acquired in the acquisition step is an image in which each pixel of the above-described skin image is represented only by the blue (B) component in the RGB color space (hereinafter, sometimes referred to as a B image). It is. For example, the B image can be acquired by setting the R value and the G value to 0 for each pixel of the skin image. When the skin image has color space information other than the RGB color space such as YCbCr, the B image can be acquired by performing a well-known color space conversion process in the acquisition process.

In the evaluation step, the state of the square plug is evaluated based on the blue component image acquired in this way.
Here, “the state of the square plug” is various states relating to the square plug present in the skin, such as size, number, distribution, etc., and can be indicated by physical quantities such as the number, area, dispersion value, etc. It can also be indicated by information such as a small amount.
There is no limitation on the evaluation method in the evaluation process. For example, the computer may evaluate the state of the horn plug of the target skin using a predetermined algorithm based on the blue component image acquired in the acquisition process. In this case, the computer can also present the evaluation results (“large”, “large and small are scattered”, “large is scattered”, “less”, “almost no”, etc.). In addition, the computer presents information obtained by applying some processing to the blue component image acquired in the acquisition process, and the person (subject, evaluator, etc.) who referred to this information shows the state of the horn plug of the target skin You may evaluate.

Since the square plugs present in the pores contain sebum and the like, the incident light is strongly reflected compared to the tissue part of the epidermis. For this reason, a square plug is represented as a local area | region with high brightness | luminance in a blue component image.
Furthermore, since the light in the blue wavelength region hardly enters the skin and is reflected by the surface of the skin, it is considered that the local region showing the horn plug shows the state of the horn plug with high accuracy.
That is, by using the blue component image, the state of the square plug can be evaluated with high accuracy. Since square plugs are clogged pores, according to such a method, clogging of pores can be evaluated with high accuracy.

Hereinafter, the above-described method will be described in more detail.
FIG. 1 is a flowchart showing a pore clogging evaluation method according to this embodiment.
The pore clogging evaluation method according to the present embodiment (hereinafter sometimes referred to as the present evaluation method) includes a step (S11), a step (S13), a step (S15), a step (S17), a step (S19), and a step. (S20) is included.

In the step (S11), a skin image (hereinafter referred to as a skin image) captured under irradiation of light having a light emission spectrum including a boundary between an ultraviolet wavelength region and a visible wavelength region within a half-value width as a main component and at least a part of a blue wavelength region. , Which is expressed as an original skin image). The boundary wavelength between the ultraviolet wavelength region and the visible light wavelength region is, for example, 390 nm. However, the Japanese Industrial Standard JIS Z 8120 stipulates that the short wavelength limit of the visible light wavelength range may be considered to be 360 nm to 400 nm, and therefore the boundary wavelength may be determined in the range of 360 nm to 400 nm. Good. The “blue wavelength region” is as described above.
The reason why light having an ultraviolet wavelength region in addition to the blue wavelength region is used is to evaluate pore clogging with high accuracy by using the presence of a substance that emits light by ultraviolet rays such as porphyrin.
It is known that there are acne bacteria that prefer sebum and produce porphyrin by metabolism in the pores, whereby porphyrin exists in the pores. Porphyrin is known as a substance that emits red light when excited by ultraviolet rays. Of course, the substance that exists in the pores and emits light by ultraviolet rays may be other than porphyrin.
Therefore, by irradiating the skin with light having an ultraviolet wavelength region in addition to the blue wavelength region, red light emitted by porphyrins present in the pores together with the square plugs in addition to the reflection component due to the square plugs of the irradiation light having the blue wavelength region By obtaining a skin image obtained by imaging components, it is possible to evaluate clogging of pores with high accuracy. In the conventional method, focusing on the characteristics of porphyrin that emits red light by ultraviolet rays, the red (R) component of the skin image is mainly analyzed. In the present embodiment, such characteristics are used supplementarily. .
Hereinafter, the light irradiated in the photographing environment of the original skin image obtained in the step (S11) may be referred to as specific light. For example, the specific light is ultraviolet light having a peak wavelength of 380 nm.

In step (S13), a red component image (hereinafter referred to as R image), a green component image (hereinafter referred to as G image), and a blue component image (B) from the skin image acquired in step (S11). Image).
The R image is an image in which each pixel of the skin image is represented by only the red (R) component in the RGB color space, and the G image is represented by only each green (G) component of the skin image. The B image is an image in which each pixel of the skin image is represented only by a blue (B) component. For example, the R image is acquired by setting the G value and the B value to 0 for each pixel of the skin image, and the G image is acquired by setting the R value and the B value to 0 for each pixel of the skin image. The B image can be obtained by setting the R value and the G value to 0 for each pixel of the skin image. In the step (S13), a well-known color space conversion process can be performed to convert the skin image into an image having RGB color space information.

Step (S15) is a step of specifying a square plug region for each of the R image, G image, and B image acquired in step (S13).
Here, the “square plug region” is a local pixel region that is presumed to represent a square plug in each image. Since the square plug is considered to have a higher reflectance than other epidermal tissues, it can be considered to be represented by a pixel having a high pixel value in each of the R image, the G image, and the B image.
Therefore, a pixel group having a pixel value greater than or equal to a predetermined threshold in each of the R image, the G image, and the B image can be specified as a square plug region. The predetermined threshold may be a different value in each of the R image, the G image, and the B image.

At this time, islands (lumps) are determined for the pixel group corresponding to the square plugs, and only a predetermined number or more of consecutive islands (lumps) are specified as the square plug region. Other pixels need not be the square plug region. As an island determination, for example, a pixel group adjacent in the vertical direction or the horizontal direction among pixels corresponding to a square plug is determined as one island (lumb), that is, a square plug region, and adjacent only diagonally. Pixels may not be determined as a cluster (island). Of course, even pixels that are adjacent only diagonally may be determined as a cluster (island).
That is, in the step (S15), regardless of whether or not it is an island, a set of individual pixels corresponding to a square plug may be specified as a “square plug region”, or an island is determined by island determination. Each pixel block may be identified as a “square plug region”.

Further, a smoothing filter such as a Gaussian filter is applied to each of the R image, the G image, and the B image, and a difference image between each obtained low-frequency image and the original image is obtained, and each difference image is obtained. A square plug region may be specified using each image obtained by binarizing each of the images. In this case, a pixel having a value corresponding to strong luminance among the binarized values may be specified as corresponding to a square plug.
Hereinafter, the plug region specified in the R image is referred to as an R plug region, the plug region specified in the G image is referred to as a G plug region, and the plug region specified in the B image is referred to as a B corner. Sometimes referred to as plug region.

  Step (S17) is a step of calculating the sizes of the R plug region, the G plug region, and the B plug region specified in step (S15). Specifically, the number of pixels specified as the corner plug region is acquired as the size of the corner plug region. At this time, when each pixel block is specified as a corner plug region, the size may be acquired for each corner plug region, or the total size of each corner plug region may be acquired as the size of the corner plug region. May be.

The step (S19) is a step of extracting size information or number of the horn plugs using the sizes of the R horn plug region, the G horn plug region, and the B horn plug region acquired in the step (S17).
In the step (S19), multiple regression equations acquired in advance by multiple regression analysis using the skin sample data of the population, and the sizes of the R plug region, the G plug region, and the B plug region are used as explanatory variables. A multiple regression equation with the size of the horn plug as the objective variable can be used. The present inventors have found that the size of an actual square plug can be estimated with high accuracy by using information of the square plug region obtained from each of the R image, the G image, and the B image.
Here, the “population skin sample data” refers to steps (S11) to (S17) for the original skin image of each subject imaged under the above-mentioned specific light irradiation with a plurality of subjects as a population. Data of each size of the R plug region, G plug region and B plug region obtained by application is included. Further, the “population skin sample data” is the size of a square plug that is visually measured from each skin image obtained by imaging the skin of each subject with a microscope at substantially the same magnification as the original skin image. Are included as ground truth data.
The explanatory variable and objective variable of the multiple regression equation, each size of the R plug region, the G plug region and the B plug region, and the size of the plug may be a size corresponding to the plug of each pore. However, it may be the total size of the square plugs of each pore.
In the step (S19), by substituting the sizes of the R square plug region, the G square plug region, and the B square plug region obtained in the step (S17) into the previously obtained multiple regression equations, The size of the horn plug as a variable can be calculated.
In the step (S19), the size of the horn plug calculated in this way may be extracted as it is as the size information of the horn plug, or based on the calculated size of the horn plug, the total area of the horn plug, the average Area or area variation information (dispersion value or the like) may be calculated, and the calculated value may be extracted as corner plug size information.

The number of square plugs can be extracted when each pixel block determined as an island in the step (S15) is specified as a square plug region. In this case, an average of the numbers of R square plug regions, G square plug regions, and B square plug regions having a size equal to or larger than a predetermined threshold may be the number of the square plugs.
In addition, a multiple regression equation obtained in advance by multiple regression analysis using the skin sample data of the population, wherein the numbers of the R plug region, the G plug region, and the B plug region are explanatory variables. The number of horn plugs can also be calculated using a multiple regression equation with number as an objective variable. Specifically, the number of square plugs as the objective variable can be calculated by substituting the numbers of the R square plug region, the G square plug region, and the B square plug region into the multiple regression equation.
In this case, the “population skin sample data” is obtained by performing steps (S11) to (S17) on the original skin image of each subject imaged under the above-described specific light irradiation with a plurality of subjects as a population. R size plug area, G square plug area, and B square plug area data obtained by applying or the number of R square plug areas, G square plug areas, and B square plug areas having a size larger than a predetermined threshold It only has to contain data. Further, the “population skin sample data” indicates the number of horn plugs visually measured from each skin image obtained by imaging the skin of each subject with a microscope at substantially the same magnification as the original skin image. As long as it is included as ground truth data.

Step (S20) is a step of evaluating the horn plug state of the skin shown in the original skin image using the size information or number of horn plugs extracted in step (S19).
The specific evaluation method in the step (S20) is not limited.
For example, the computer may output the size information or number of the horn plugs extracted in step (S19), and a person who refers to the output may evaluate the state of the horn plugs. At this time, the computer may further output the size information or the number of the square plugs used as the evaluation criterion, or such information may be output in advance on a print medium or the like. The size information or the number of horn plugs used as an evaluation criterion can be acquired from the average skin or ideal skin for each age group by the same method as in this embodiment.
In addition, the computer compares the size information or number of the horn plugs extracted in step (S19) with the size information or number of the horn plugs used as the above-mentioned evaluation criteria, or other comparison target data, Evaluation results can also be output.

The pore clogging evaluation method according to the present embodiment described above can be executed by, for example, a pore clogging evaluation apparatus illustrated in FIGS. 2 and 3 (hereinafter, sometimes abbreviated as an evaluation apparatus).
FIG. 2 is a diagram conceptually illustrating a hardware configuration example of the pore clogging evaluation apparatus 10 according to the present embodiment.
The evaluation device 10 is a so-called computer, and includes, for example, a CPU (Central Processing Unit) 11, a memory 12, an input / output interface (I / F) 13, a communication unit 14, and the like that are connected to each other via a bus. The number of hardware elements forming the evaluation device 10 is not limited, and these hardware elements can be collectively referred to as an information processing circuit. Moreover, the evaluation apparatus 10 may include hardware elements not illustrated in FIG. 2, and the hardware configuration is not limited.

The CPU 11 may be configured by an application specific integrated circuit (ASIC), a DSP (Digital Signal Processor), a GPU (Graphics Processing Unit), or the like in addition to a general CPU.
The memory 12 is a RAM (Random Access Memory), a ROM (Read Only Memory), or an auxiliary storage device (such as a hard disk).
The input / output I / F 13 can be connected to user interface devices such as the output device 15 and the input device 16. The output device 15 is at least one of a device such as an LCD (Liquid Crystal Display) or a CRT (Cathode Ray Tube) display, a device that displays a screen corresponding to drawing data processed by the CPU 11, a printing device, and the like. The input device 16 is a device that receives an input of a user operation such as a keyboard and a mouse. The output device 15 and the input device 16 may be integrated and realized as a touch panel.
The communication unit 14 performs communication via a communication network with other computers, exchange of signals with other devices, and the like. A portable recording medium or the like can be connected to the communication unit 14. The communication unit 14 may be connected to a digital camera (not shown) that captures the above-described original skin image.
Further, the evaluation device 10 may be a device incorporating the digital camera.

FIG. 3 is a diagram conceptually illustrating a processing configuration example of the pore clogging evaluation apparatus 10 according to the present embodiment.
The evaluation apparatus 10 includes an acquisition unit 21, a specification unit 22, a calculation unit 23, an extraction unit 24, and an evaluation unit 25. These are software elements, and are realized by, for example, the CPU 11 executing an evaluation program 19 stored in the memory 12. This program is installed from, for example, a portable recording medium such as a CD (Compact Disc) or a memory card or another computer on the network via the input / output I / F 13 or the communication unit 14, and stored in the memory 12. May be.

The acquisition part 21 performs a process (S11).
The acquisition unit 21 can acquire an original skin image from an external computer, device, portable recording medium, or the like via the input / output I / F 13 or the communication unit 14. Moreover, when the acquisition part 21 is implement | achieved as a camera, the acquisition part 21 may image and produce | generate the original skin image.
Furthermore, the acquisition part 21 performs a process (S13). The method for acquiring the R image, the G image, and the B image from the original skin image is as described above.

The specifying unit 22 performs the process (S15). The definition of the R plug region, the G plug region, and the B plug region and the method for acquiring them are as described above.
The calculating unit 23 executes the process (S17). The sizes of the R plug region, the G plug region, and the B plug region and the calculation methods thereof are as described above.
The extraction part 24 performs a process (S19). The method for extracting the size information or number of horn plugs is as described above. The extraction unit 24 may hold in advance a multiple regression equation for calculating the size of the square plug.

Evaluation part 25 performs a process (S20). That is, the evaluation unit 25 generates and outputs evaluation result information of the skin plug state in the original skin image. The specific evaluation method by the evaluation unit 25 and the specific contents of the output evaluation result are not limited.
For example, the evaluation unit 25 may output, as an evaluation result, one or both of the size information and the number of square plugs extracted by the extraction unit 24. In addition, the evaluation unit 25 may output the size information or the number of square plugs used as an evaluation reference so that the evaluation result can be compared with the evaluation result. In this case, the evaluation part 25 should just hold | maintain the magnitude | size information or number of the square plug used as evaluation criteria previously.
Further, the evaluation unit 25 compares the information of the evaluation criteria with the information extracted by the extraction unit 24, and outputs information such as good, average, slightly bad, and bad as the evaluation result of the horn plug state. You can also In this case, the evaluation unit 25 holds in advance threshold groups for leveling by age group, sex, etc., and selects and selects a threshold corresponding to the age, sex, etc. of the subject from among them. You may produce | generate and output the evaluation result of a test subject's horn plug state using a threshold value.

[Modification]
The present invention is not limited to the above-described embodiment, and the above-described embodiment allows various modifications and improvements as long as the object of the present invention is achieved.

  For example, the above-described evaluation method may not include the step (S11). That is, this evaluation method may be started from acquiring an R image, a G image, and a B image acquired from an original skin image. Similarly, the acquisition unit 21 of the evaluation apparatus 10 does not acquire the original skin image, and outputs the R image, the G image, and the B image from an external computer, device, portable recording medium, or the like, or the communication unit 14. You may get through.

In addition, as will be described in detail in Examples, the present inventors have found that only the information of the B horn plug region of the B image shows a high correlation with the size of the horn plug.
Therefore, this evaluation method may not use one or both of the R image and the G image. That is, in the step (S13), only the B image may be acquired, or either the B image and either the R image or the G image may be acquired. Accordingly, in the step (S15), only the B plug region may be specified, or either the B plug region and either the R plug region or the G plug region may be specified. In the step (S17), only the size of the B square plug region may be calculated, or the size of the B square plug region and either the R square plug region or the G square plug region may be calculated. . In the step (S19), only the size of the B square plug region calculated in the step (S17), or the size of either the B square plug region and the size of either the R square plug region or the G square plug region, What is necessary is just to extract the size information or number of plugs. In this case, in the step (S19), a single regression equation with the size of the plug as an objective variable and the size of the B plug region as an explanatory variable, or the size of the B plug region and R A multiple regression equation with the size of either the square plug region or the G square plug region as an explanatory variable can be used.
In this case, the acquisition unit 21, the identification unit 22, the calculation unit 23, and the extraction unit 24 of the evaluation device 10 that performs these steps may execute the processing content of each step modified as described above.
Moreover, although the clogging clogging evaluation method was illustrated in the above-mentioned embodiment, a square plug information extraction method can be illustrated as this embodiment. This corner plug information extraction method includes an acquisition step of acquiring a blue component image from a skin image captured under light irradiation including a part of a blue wavelength region, and a corner plug region in the acquired blue component image. What is necessary is just to include the specific process to identify, and the extraction process which extracts the number or magnitude | size information of a square plug using the specified square plug area | region. The light including a part of the blue wavelength region includes an emission spectrum component whose power at 390 nm is 10% or more of the maximum value. In the extraction step, the light is preliminarily analyzed by regression analysis using population skin sample data. Using the regression equation (single regression equation or multiple regression equation) obtained and using the regression equation with the size of the identified plug region as the explanatory variable and the plug size as the objective variable, the size of the plug Extract information.

  Part or all of the above-described embodiments and modifications may be specified as follows. However, the above-described embodiment and modification examples are not limited to the following description.

<1> an acquisition step of acquiring a blue component image from a skin image captured under irradiation of light including at least a part of the blue wavelength region;
An evaluation step for evaluating the state of the horn plug based on the acquired blue component image;
A method for evaluating clogged pores.

<2> The evaluation step is:
A specifying step for specifying a first square plug region in the acquired blue component image;
An extraction step of extracting the number or size information of the horn plugs using the identified first horn plug region;
<1> The pore clogging evaluation method according to <1>.
<3> The evaluation step is:
A calculation step of calculating a size of the identified first horn plug region;
In the extraction step, size information of the square plug is extracted using the calculated size.
The pore clogging evaluation method according to <2>.
<4> The light includes an emission spectrum component whose power at 390 nm is 10% or more of the maximum value,
In the acquisition step, one or both of a red component image or a green component image is further acquired from the skin image,
In the specifying step, one or both of specifying a second horn plug region of the acquired red component image or specifying a third triangular plug region of the acquired green component image is further executed,
In the calculation step, the size of one or both of the second square plug region or the third triangular plug region is further calculated,
In the extraction step, the size information of the square plug is extracted using the size of one or both of the second square plug region or the third triangular plug region and the size of the first square plug region.
The pore clogging evaluation method according to <3>.
<5> In the extraction step, the total area, average area, or area variation information of the square plug is extracted as the size information of the square plug.
The pore clogging evaluation method according to <4>.
<6> In the extraction step, a multiple regression equation acquired in advance by multiple regression analysis using skin sample data of a population, wherein the first horn plug region, the second horn plug region, and the third triangle plug Extract size information of horn plugs using a multiple regression equation where each size of the region is an explanatory variable and each horn plug size is an objective variable.
The pore clogging evaluation method according to <4> or <5>.
<7> Acquisition means for acquiring a blue component image from a skin image captured under irradiation of light including at least a part of the blue wavelength region;
Evaluation means for generating evaluation result information of the state of the horn plug based on the acquired blue component image;
An apparatus for evaluating clogged pores.
<8> A specifying unit that specifies the first horn plug region in the blue component image acquired by the acquiring unit;
Extraction means for extracting the number or size information of the horn plugs using the first horn plug region specified by the specifying means;
The pore clogging evaluation apparatus according to <7>, further comprising:
<9> Calculation means for calculating the size of the first square plug region specified by the specifying means,
Further comprising
The extraction means extracts size information of the horn plug using the size calculated by the calculation means.
The pore clogging evaluation apparatus according to <8>.
<10> The light includes an emission spectrum component whose power at 390 nm is 10% or more of the maximum value,
The acquisition means further acquires one or both of a red component image and a green component image from the skin image,
The specifying means specifies either the second square plug region of the red component image acquired by the acquisition means, or specifies the first triangular plug region of the green component image acquired by the acquisition means or Do both further,
The calculation means further calculates the size of one or both of the second square plug region or the third triangular plug region,
The extraction means extracts the size information of the square plug using the size of one or both of the second square plug region or the third triangular plug region and the size of the first square plug region.
<9> The clogging evaluation apparatus according to <9>.
<11> The extraction means extracts the total area, average area, or area variation information of the square plugs as square plug size information.
<10> The clogging evaluation apparatus according to <10>.
<12> The extraction means is a multiple regression equation acquired in advance by multiple regression analysis using population skin sample data, wherein the first square plug region, the second square plug region, and the third triangular plug Extract size information of horn plugs using a multiple regression equation where each size of the region is an explanatory variable and each horn plug size is an objective variable.
<10> or <11> The pore clogging evaluation apparatus according to <11>.
<13> An acquisition step of acquiring a blue component image from a skin image captured under light irradiation including a part of the blue wavelength region;
A specifying step of specifying a corner plug region in the obtained blue component image;
An extraction step of extracting information on the number or size of horn plugs using the specified horn plug region;
A plug extraction method including
The light includes an emission spectrum component whose power at 390 nm is 10% or more of the maximum value,
In the extraction step, a regression equation obtained in advance by a regression analysis using the skin sample data of the population, the regression including the size of the identified plug region as an explanatory variable and the plug size as an objective variable Extract the size information of the horn plug using the formula,
Square plug information extraction method.

  Examples will be given below to explain the above-mentioned contents in more detail. However, the description of the following examples does not limit the above-described contents at all.

In the present embodiment, the skin of each subject is a digital camera in an environment where a group of five subjects is irradiated with specific light having a wavelength spectrum with an emission peak wavelength of 380 nm and a blue component wavelength of about 450 nm. The original skin images for five persons were acquired by imaging each of (see FIG. 4A).
On the other hand, the same location as the imaging location of the original skin image in each subject's skin was each image | photographed with the microscope, and the ground truth skin image for 5 persons was acquired (refer FIG.4 (b)).
FIG. 4A is a diagram illustrating an example of an original skin image, and FIG. 4B is a diagram illustrating an example of a ground truth skin image.

  The person in charge identified 7 to 8 horn plugs in the ground truth skin image of each subject, and measured the long diameter of each horn plug. As a result, the major axis for each of the 38 square plugs for each of the five skins was acquired as ground truth data.

The R image, G image, and B image were respectively acquired from the original skin images for the five persons acquired as described above, and the following processing was performed on each of the R image, the G image, and the B image. . A smoothing filter was applied to the original image, a difference image between the obtained smoothed image and the original image was acquired, and the difference image was binarized. Thereby, a binarized R image, a binarized G image, and a binarized B image were acquired for the R image, the G image, and the B image.
The plug region was specified for the acquired binarized R image, binarized G image, and binarized B image, and the R plug region, the G plug region, and the B plug region were specified. . In this embodiment, island identification is performed in specifying the square plug region, and among the pixels corresponding to the square plug, a pixel group adjacent in the vertical direction or the horizontal direction is specified as one square plug region.
Subsequently, the number of pixels for each square plug region was calculated as the area of the square plug region.
As a result, regarding the skin for five persons, the number of R-plug regions and the area for each R-plug region, the number of G-plug regions and the area for each G-plug region, and the number of B-plug regions and B-horn The area for each plug region was acquired, and these data and the above-mentioned ground truth data were combined into skin sample data for a population of five people.

By thus obtained skin sample data multiple regression analysis and multiple regression equation and the coefficient of determination R ^2, such as can be obtained or less, the result is significant at a significance level less than 1%.
Square plug major axis = 100.4699−0.25923 × (R square plug region) −0.04209 × (G square plug region) +1.514402 (B square plug region)
Determination coefficient R ² = 0.550062
By using such a multiple regression equation, the number of horn plugs (eg, a predetermined size or more), the long diameter of each horn plug, and the average long diameter of the horn plugs without visual measurement from the original skin image of the subject. It is possible to estimate the size information or number of square plugs with high accuracy, such as the variance of the long diameter of the square plugs.
In the above-described embodiment, such a multiple regression equation may be acquired and used in advance.
In this embodiment, the long diameter of the square plug is measured as the ground truth data, but the area of the square plug may be measured. Further, although the number of pixels of the square plug region is acquired as the area, the number of pixels arranged in the major axis of the square plug region may be acquired as the size of the square plug region.

Further, in this embodiment, the skin sample data as described above is used to determine the area of the horn plug region and the actual size of the horn plug region for each of the R plug region, the G plug region, and the B plug region. The relationship was investigated.
5 is a graph showing the relationship between the area of the R plug region and the long diameter of the plug, and FIG. 6 is a graph showing the relationship between the area of the G plug region and the long diameter of the plug. These are graphs showing the relationship between the area of the B plug region and the major axis of the plug.
According to the determination coefficient R2 shown in FIG. 5, FIG. 6, and FIG. 7, from the form using all information of the R plug region, the G plug region, and the B plug region (determination coefficient R ² = 0.550062). However, it has been proved that the size information or the number of the horn plugs can be estimated with sufficiently high accuracy only by the information of the B horn plug region of the B image.
Although not illustrated as an example, the size information or the number of horn plugs is estimated with sufficiently high accuracy from the information of the B horn plug region and the information of either the R horn plug region or the G horn plug region. It has been demonstrated that it can be done.

Furthermore, as a comparative example, a method for extracting the size of the horn plug region from the above-mentioned five original skin images (RGB images) without acquiring the R image, the G image, and the B image was investigated.
In this comparative example, a smoothing filter is applied to each of five original skin images (RGB images), and a difference image between the obtained smoothed image and the original image is acquired. On the other hand, binarization was performed. Then, in the image obtained by the binarization, a corner plug region (expressed as an RGB corner plug region) was specified, and the area of the RGB corner plug region was calculated.
FIG. 8 is a graph showing the relationship between the area of the RGB square plug region and the long diameter of the square plug as a comparative example.
As shown in FIG. 8, in this comparative example, the correlation between the area of the horn plug region and the major axis of the horn plug was lower than that of the method of the above-described embodiment. That is, this comparative example proves that the method of the present embodiment can estimate the size information or the number of square plugs with higher accuracy.

10 Pore clogging evaluation device (Evaluation device)
11 CPU
12 Memory 13 Input / output I / F
14 Communication unit 15 Output device 16 Input device 19 Evaluation program 21 Acquisition unit 22 Identification unit 23 Calculation unit 24 Extraction unit 25 Evaluation unit

Claims (8)

An acquisition step of acquiring a blue component image from a skin image captured under irradiation of light including at least a part of the blue wavelength region;
An evaluation step for evaluating the state of the horn plug based on the acquired blue component image;
A method for evaluating clogged pores. The evaluation step includes
A specifying step for specifying a first square plug region in the acquired blue component image;
An extraction step of extracting the number or size information of the horn plugs using the identified first horn plug region;
The pore clogging evaluation method according to claim 1, comprising: The evaluation step includes
A calculation step of calculating a size of the identified first horn plug region;
In the extraction step, size information of the square plug is extracted using the calculated size.
The pore clogging evaluation method according to claim 2. The light includes an emission spectrum component whose power at 390 nm is 10% or more of the maximum value,
In the acquisition step, one or both of a red component image or a green component image is further acquired from the skin image,
In the specifying step, one or both of specifying a second horn plug region of the acquired red component image or specifying a third triangular plug region of the acquired green component image is further executed,
In the calculation step, the size of one or both of the second square plug region or the third triangular plug region is further calculated,
In the extraction step, the size information of the square plug is extracted using the size of one or both of the second square plug region or the third triangular plug region and the size of the first square plug region.
The pore clogging evaluation method according to claim 3. In the extraction step, the total area of the square plug, the average area or the variation information of the area is extracted as the size information of the square plug.
The pore clogging evaluation method according to claim 4. In the extraction step, a multiple regression equation obtained in advance by multiple regression analysis using population skin sample data, each of the first horn plug region, the second horn plug region, and the triangular plug region Extract the size information of the angle plug using a multiple regression equation with the size as the explanatory variable and the size of the individual angle plug as the objective variable.
The pore clogging evaluation method according to claim 4 or 5. An acquisition means for acquiring a blue component image from a skin image captured under irradiation of light including at least a part of the blue wavelength region;
Evaluation means for generating evaluation result information of the state of the horn plug based on the acquired blue component image;
An apparatus for evaluating clogged pores. An acquisition step of acquiring a blue component image from a skin image captured under irradiation of light including a part of the blue wavelength region;
A specifying step of specifying a corner plug region in the obtained blue component image;
An extraction step of extracting information on the number or size of horn plugs using the specified horn plug region;
A plug extraction method including
The light includes an emission spectrum component whose power at 390 nm is 10% or more of the maximum value,
In the extraction step, a regression equation obtained in advance by a regression analysis using the skin sample data of the population, the regression including the size of the identified plug region as an explanatory variable and the plug size as an objective variable Extract the size information of the horn plug using the formula,
Square plug information extraction method.