JP2017192767A - Image analysis method, image analysis device, and image analysis program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To perform accurately, analysis of a skin or the like using a photographic image of a subject.SOLUTION: There is provided an image analysis method for using an image of a subject photographed in a prescribed wavelength region for performing analysis of a skin of the subject, comprising: a mesh arrangement step for setting a shape of meshes, based on a reference point obtained from the image, then arranging plural meshes, based on the set shape of meshes; a statistic amount calculation step for calculating a statistic amount for every mesh obtained in the mesh arrangement step; an image analysis step for analyzing a skin for every mesh, based on the statistic amount obtained in the statistic amount calculation step; and an image generation step for generating an image of the analysis result obtained in the image analysis step.SELECTED DRAWING: Figure 3

Description

  The present invention relates to an image analysis method, an image analysis apparatus, and an image analysis program, and in particular, an image analysis method, an image analysis apparatus, and an image analysis program for analyzing skin and the like with high accuracy using a photographed image of a subject. About.

  2. Description of the Related Art Conventionally, there is a technique for analyzing a subject's skin (skin) state using an image of the subject. For example, in the past, there has been a technology that takes two near-infrared spectral images using a near-infrared camera to visualize changes in the moisture on the face, and visualizes the moisturizing liquid applied to the face surface from the taken images. It is disclosed (for example, Non-Patent Document 1). In recent years, a technique for discriminating the amount of skin moisture using near infrared has been disclosed (see, for example, Patent Document 1).

  Furthermore, in recent years, a face image before and after applying a skin external preparation to a subject is acquired, a difference amount between the acquired images before and after is acquired for each of a plurality of different near infrared regions, and the subject is based on the acquired difference amount. There is a technique for analyzing the skin of the above (for example, see Patent Document 2).

JP 2010-25622 A Japanese Patent No. 483590

Hiroaki Iwasaki et al., “Visualization of moisture change in face using two near-infrared spectral images”, The Optical Society of Japan (Applied Physics Society), Optics Japan 2005 Tokyo, November 23-25, 2005

  According to the conventional method as described above, the skin state of each subject can be analyzed. However, since subjects have different face shapes and sizes, for example, statistical analysis and evaluation on the same area of the skin targeting a plurality of subjects cannot be performed. Therefore, it has been impossible to analyze skin and the like with high accuracy using a photographed image of the subject.

  The present invention has been made in view of the above-described problems, and provides an image analysis method, an image analysis apparatus, and an image analysis program for performing analysis of skin and the like with high accuracy using a photographed image of a subject. For the purpose.

  In order to solve the above-described problems, the present invention employs means for solving the problems having the following characteristics.

The invention described in claim 1 is an image analysis method for analyzing a skin of a face of a subject using a face image of the subject photographed in a predetermined wavelength region.
An area setting step for setting the vertical and horizontal lengths of a mesh area in which a plurality of meshes are arranged based on at least three reference points including at least the corners of the eye obtained from the face image;
The vertical length of the mesh area is calculated by dividing the vertical width of the mesh area by the vertical number of meshes, and the horizontal width of the mesh area is calculated by dividing the horizontal width of the mesh area by the horizontal number of meshes. A mesh placement step of calculating a thickness and placing a plurality of polygonal meshes of the same type of the calculated size;
A statistic calculation step for calculating a statistic for each mesh obtained by the mesh placement step;
An image analysis step for analyzing the skin for each mesh based on the statistic obtained by the statistic calculation step;
An image generation step of generating an image of the analysis result obtained by the image analysis step.

  According to the present invention, analysis of skin or the like can be performed with high accuracy using a photographed image of a subject.

It is a figure which shows an example of a function structure of the image analysis apparatus in 1st Embodiment. It is a figure which shows an example of the hardware constitutions which can implement | achieve the image analysis process in 1st Embodiment. It is a flowchart which shows an example of the image-analysis processing procedure in 1st Embodiment. It is a figure for demonstrating the difference in a mesh shape. FIG. 3 is a diagram for explaining an example of inputting statistics to a map diagram according to the first embodiment (No. 1); FIG. 6 is a diagram (part 2) for explaining an example of inputting statistics to a map diagram in the first embodiment. It is a figure which shows an example of the analysis content corresponding to a statistic. It is a figure which shows an example of a moisture map. It is a figure which shows an example of a dispersion | distribution map. It is a figure which shows an example of the level distribution corresponding to a skin state. It is a figure which shows an example of a function structure of the image analysis apparatus in 2nd Embodiment. It is a flowchart which shows an example of the image-analysis processing procedure in 2nd Embodiment. It is a figure for demonstrating an example of the acquisition method of an emphasized image. It is a figure which shows an example of the analysis result in 2nd Embodiment.

<About the present invention>
In the present invention, when the skin area of the subject included in the image is divided into meshes, the size of the mesh can be changed based on the face shape of each subject, the size of the face in the image, the resolution, and the like. It can be so. That is, in the present invention, for example, the number of meshes divided for each subject is made the same, and the size of the mesh is changed based on the relative positional relationship. Further, according to the present invention, a statistic for each mesh is calculated and used as one of analysis (evaluation) indices.

  Therefore, in the present invention, the area to be compared in the whole face is subdivided in advance, and the skin statistics with higher accuracy are realized by comparing the statistics of each area based on the relative position of the subdivided area for each subject. can do.

  Hereinafter, embodiments in which an image analysis method, an image analysis apparatus, and an image analysis program according to the present invention are suitably implemented will be described with reference to the drawings.

<First embodiment: image analysis apparatus: schematic configuration example>
First, a functional configuration example of the image analysis apparatus according to the first embodiment will be described with reference to the drawings. FIG. 1 is a diagram illustrating an example of a functional configuration of the image analysis apparatus according to the first embodiment. An image analysis apparatus 10 shown in FIG. 1 includes an input unit 11, an output unit 12, a storage unit 13, an image acquisition unit 14, a correction unit 15, a mesh arrangement unit 16, a statistic calculation unit 17, and an image. The analysis unit 18, the image generation unit 19, the transmission / reception unit 20, and the control unit 21 are included.

  The input unit 11 receives input of various instructions such as an image acquisition instruction, a correction instruction, a mesh arrangement instruction, a statistic calculation instruction, an image generation instruction, and a transmission / reception instruction from a user or the like. Note that the input unit 11 includes, for example, a keyboard and a pointing device such as a mouse. The input unit 11 also has a function of inputting a skin image of a subject photographed by a preset imaging unit such as a digital camera or a near infrared camera.

  The output unit 12 displays / outputs the content input by the input unit 11 and the content executed based on the input content. The output unit 12 includes a display, a speaker, and the like. Further, the output unit 12 may have a function of a printer or the like. In that case, the image analysis result or the like can be printed on a print medium such as paper and provided to the user or the subject.

  Note that the input unit 11 and the output unit 12 may be integrated input / output units such as a touch panel. In this case, the input unit 11 and the output unit 12 may be predetermined using a user's finger or a pen-type input device. Input can be made by touching the position.

  The storage unit 13 stores various data such as a captured image obtained by the captured image acquisition unit 14, a corrected image obtained by the correcting unit, mesh arrangement information, statistics, image analysis results, image generation information, and setting information. . Further, the storage means 13 can read out various data stored as necessary.

  The captured image acquisition unit 14 acquires a face image of the subject captured by an imaging unit such as a camera. The captured image acquisition unit 14 can acquire, for example, a visible image, a near infrared image, an infrared image, an ultraviolet image, or the like based on a preset wavelength region, but is not limited thereto. Further, the acquired image may be a gray scale luminance image or the like in the case of black and white, and may be an RGB luminance image or the like in the case of color.

  In addition, the photographed image acquisition unit 14 can set, for example, the type, position, number, and the like of the light source to be used according to the content of the photographed image when photographing the subject's face. The image acquired by the captured image acquisition unit 14 is accumulated in the storage unit 13.

  The correction unit 15 performs correction processing such as color correction, contrast correction, and brightness correction on the image obtained by the image acquisition unit 14. The correction process may be performed as necessary. For example, an operator (user) who operates the image analysis apparatus 10 may set in advance whether or not to perform the correction process.

  The mesh arrangement means 16 sets a reference point for dividing the mesh into the skin area of the subject included in the image. Specifically, for example, when mesh division is performed on the face image of the subject, the mesh arrangement unit 16 sets three or more points with reference to a portion (feature point) with less movement on the face. In the first embodiment, by setting three or more points, the vertical and horizontal ratio of the mesh can be calculated. Examples of the reference point include the corner of the eye, the head of the eye, the tip of the nose, the tip of the chin, the center of the upper lip, and the ear hole, but are not limited thereto, and may be, for example, the center of the eye. In the first embodiment, since the face image to be shot may be a front image or a side image, the setting of the reference point can be changed according to the face area included in the shot image.

  Further, the mesh placement unit 16 calculates the size of the mesh based on the set reference point. For example, the mesh arrangement means 16 makes the mesh rectangular, and when the vertical number and horizontal number of meshes based on the reference point are set in advance, the vertical width obtained from the plurality of reference points is set to the vertical length of the mesh. By dividing by the number, the vertical length of the mesh is calculated, and by dividing the horizontal width obtained from a plurality of reference points by the horizontal number of the mesh, the horizontal length of the mesh is calculated. . Thereby, the vertical and horizontal lengths of the mesh are set.

  Further, the mesh placement unit 16 places a plurality of meshes (mesh groups) in association with the positions of the reference points based on the calculated mesh size. Note that the same number of meshes is arranged with respect to the same face direction of each subject. That is, in the first embodiment, for example, the relative position and number of meshes in a specific analysis region such as the forehead and cheek are fixed. The mesh shape may be a polygon selected in advance, such as a rectangle or a triangle, but is not limited to this.

  The statistic calculation means 17 calculates the statistic of each mesh. Examples of the statistics include, but are not limited to, an average value of luminance for each pixel included in each mesh, a standard deviation of luminance indicating variation in luminance with other pixels, or both. is not. Moreover, the statistic calculation means 17 calculates a statistical level from the histogram of the statistic of every test subject or all the test subjects.

  The image analysis means 18 performs image analysis in the subject group based on the statistics. Specifically, the image analysis unit 18 classifies each mesh into a predetermined number of levels (for example, five levels) using percentile values in a predetermined arbitrary mesh group. For example, the image analysis unit 18 analyzes, for example, the amount of water (the amount of water is high or low), skin variation, skin roughness, roughness, and the like based on the statistics of each mesh.

  Note that the image analysis means 18 may be performed on the entire skin of the face where the image analysis portion is taken, or only for a predetermined region designated by the user or the like using the input means 11 or the like. You may go. Further, the designated area may be one or plural. Further, the image analyzed by the image analysis unit 18 is a skin image such as a cheek or forehead, and may be an image of an arm, a hand, a leg, or the like. Furthermore, the image analysis means 18 can also analyze hair.

  Further, the image analysis means 18 does not perform analysis for each mesh but, for example, collects a plurality of meshes for each characteristic site (for example, heel, cheek, mouth) as one region, and determines the skin from the mesh statistics for that region. Analysis may be performed.

  The image generation unit 19 generates a predetermined image for displaying a statistic amount to a user, a subject, or the like. Specifically, the image generation unit 19 generates, for example, an image of a face numerical value table (map) in which statistics are input to each mesh. In addition, the image generation unit 19 defines a predetermined level for, for example, an average value or a standard deviation, and generates an image that is color-coded based on the level. In the first embodiment, the present invention is not limited to color classification. For example, pattern classification may be performed based on a predetermined level, or a color and a pattern may be combined.

  Further, the image generation means 19 may give a color or a pattern only to the outer frame of the mesh, for example, instead of giving a color or a pattern to the entire mesh area. Further, the color coding or the like may be translucent or not transparent so that a face image as a background can be seen. Moreover, the image generation means 19 can generate | occur | produce the image matched with the statistical level using the histogram etc. of the statistics of every test subject or all the test subjects. Thereby, in 1st Embodiment, a statistical result etc. can be provided to a user, a test subject, etc. more clearly.

  In addition, when displaying the generated image on the screen of the output unit 12, the image generation unit 19 displays the image after performing processing such as enlargement of a predetermined region or luminance inversion for easy viewing by the user or the subject. You can also. The image generation unit 19 performs a process such as pseudo color generation for an area set by the user in the image.

  The transmission / reception means 20 is a communication means for transmitting / receiving data to / from an external device via a communication network represented by, for example, the Internet or a LAN (Local Area Network). The transmission / reception means 20 can receive face image data, subject information, and the like that are already stored in the external device and the results analyzed by the image analysis device 10 are transmitted to the external device via the communication network. It can also be sent.

  The control unit 21 controls the entire components of the image analysis apparatus 10. Specifically, the control unit 21 performs, for example, an image acquisition process, a correction process, a mesh arrangement process, a statistic calculation process, an image analysis process, an image generation process, and the like based on an instruction from the input unit 11 by a user or the like. Perform each control.

<Image Analysis Device 10: Hardware Configuration>
Here, in the image analysis apparatus 10 described above, an execution program (image analysis program) that can cause a computer to execute each function is generated, and the execution program is installed in, for example, a general-purpose personal computer or server. The image analysis processing in the present invention can be realized. Here, an example of a hardware configuration of a computer capable of realizing the image analysis processing in the first embodiment will be described with reference to the drawings.

  FIG. 2 is a diagram illustrating an example of a hardware configuration capable of realizing the image analysis processing according to the first embodiment. 2 includes an input device 31, an output device 32, a drive device 33, an auxiliary storage device 34, a memory 35, a CPU (Central Processing Unit) 36 for performing various controls, and a network connection device 37. Are connected to each other via a system bus B.

  The input device 31 has a pointing device such as a keyboard and a mouse operated by a user or the like, and inputs various operation signals such as execution of a program from the user or the like. The input device 31 can also input an image obtained from an external device connected to the network connection device 37 or the like via the communication network or an image obtained by the image acquisition unit 14.

  The output device 32 has a display for displaying various windows and data necessary for operating the computer main body for performing the processing according to the present invention, and displays the program execution progress and results by the control program of the CPU 36. can do. Further, the output device 32 can print the above-described processing result or the like on a print medium such as paper and present it to a user, a subject, or the like.

  Here, the execution program installed in the computer main body in the present invention is provided by, for example, a portable recording medium 38 such as a USB (Universal Serial Bus) memory, a CD-ROM, or a DVD. The recording medium 38 on which the program is recorded can be set in the drive device 33, and the execution program included in the recording medium 38 is installed in the auxiliary storage device 34 from the recording medium 38 via the drive device 33.

  The auxiliary storage device 34 is a storage means such as a hard disk, and can store an execution program in the present invention, a control program provided in a computer, and the like, and can perform input / output as necessary.

  The memory 35 stores an execution program read from the auxiliary storage device 34 by the CPU 36. Note that the memory 35 includes a ROM (Read Only Memory), a RAM (Random Access Memory), and the like.

  The CPU 36 controls processing of the entire computer, such as various operations and input / output of data with each hardware component, based on a control program such as an OS (Operating System) and an execution program stored in the memory 35. Each processing can be realized. Various information necessary during the execution of the program can be acquired from the auxiliary storage device 34, and the execution result and the like can also be stored.

  The network connection device 37 acquires an execution program from another terminal connected to the communication network by connecting to a communication network or the like, or an execution result obtained by executing the program or an execution in the present invention The program itself can be provided to other devices.

  The network connection device 37 can also acquire various data such as a captured image and subject information stored in an external device connected to the communication network.

  With the hardware configuration as described above, the image analysis processing in the present invention can be executed. Further, by installing the program, the image analysis processing according to the present invention can be easily realized by a general-purpose personal computer or the like.

<Image Analysis Processing Procedure in First Embodiment>
Next, an image analysis processing procedure in the first embodiment using the image analysis apparatus 10 and the image analysis program described above will be described. FIG. 3 is a flowchart illustrating an example of an image analysis processing procedure in the first embodiment.

  In the image analysis process shown in FIG. 3, first, a captured image is acquired (S01). Note that the acquired image includes, for example, a visible region captured image (visible image), a near infrared region captured image (near infrared image), an infrared region captured image (infrared image), and an ultraviolet image captured image (ultraviolet image). ) Etc., but is not limited to this.

  Next, in the image analysis process, at least one correction process, such as color correction, contrast correction, and brightness correction, is performed on the image acquired in S01 (S02). The process of S02 corrects an image so that a plurality of images can be analyzed under the same conditions, for example. In the first embodiment, the process of S02 need not be performed.

  Next, the image analysis process sets a reference point for the corrected image (S03). Specifically, for example, a part of the face with less movement is set as a reference point. In addition, it is preferable that three or more reference points are set at different points rather than on the same straight line so that the vertical / horizontal ratio can be calculated. Examples of the reference point include, but are not limited to, both the corners of the eyes, the heads of the eyes, the tip of the nose, the tip of the chin, the center of the upper lip, and the ear hole.

  Next, the image analysis process calculates a preset mesh size based on the reference point (S04). For example, in the case of arranging a rectangular mesh, a predetermined number (for example, 10 to 15) of meshes is arranged with respect to the width of both eye corners. The length is set. Further, in the first embodiment, a vertical mesh of one mesh is arranged according to the number of meshes arranged when a predetermined number (for example, 10 to 15) of meshes is arranged with respect to the vertical width from the corner of the eye to the center of the upper lip. The length is set. Thereby, the vertical and horizontal lengths of the mesh for each subject are determined. Note that the number of mesh arrangements can be set according to, for example, the number of pixels of the image or the ratio of the face area included in the image, but is not limited thereto. For example, in the first embodiment, when the mesh is a triangle, the size of the mesh may be set based on a similar shape of a triangle having three reference points as vertices.

  Next, in the image analysis process, a mesh group is arranged around a predetermined reference position (S05). In the process of S05, the same quantity of mesh is arranged for each subject. In the process of S05, the relative position of the mesh in a specific analysis region such as the forehead or cheek can be fixed.

  Next, the image analysis process calculates a statistic in each mesh (S06). Examples of the statistics include, but are not limited to, an average value of brightness and a standard deviation. In the first embodiment, for example, when there are moles, spots, scratches, or the like existing in the mesh, the average value or standard deviation of the luminance is obtained using an area other than the area by image analysis or the like. it can.

  Next, an image analysis process produces | generates the screen which displays the statistics mentioned above. Specifically, in the first embodiment, the statistic for each mesh obtained in the process of S06 is input to a map diagram of a mesh corresponding to a predetermined position of a preset face image (S07).

  Next, in the image analysis process, for example, a level is statistically calculated using a histogram of statistics for each subject or all subjects (S08), and the mesh placed on the face image at the calculated arbitrary level is calculated. Are displayed in different colors (S09). Thereby, a user or a test subject can grasp | ascertain easily the mode of the brightness | luminance for every mesh. The image displayed on the screen in the first embodiment is not limited to this, and other various screens can be displayed.

<Analysis method according to the acquired image>
In the first embodiment, various types of analysis can be performed according to the acquired image. For example, in the case of a visible image (for example, a wavelength region of about 380 nm to about 750 nm), a distribution of the entire skin color or a portion different from the average skin color is extracted from the color image, and the distribution or the portion is extracted. Based on the analysis of the skin.

  Further, in the case of a visible image, it is possible to analyze a color that causes cosmetic problems such as smoothness (cleanliness), spots, and redness of the skin as seen from the color. Further, in the case of a visible image, the obtained color image can be RGB-decomposed to obtain the brightness gradation, so that the analysis for each mesh described above can be performed and the above-described analysis method can be applied.

  Further, in the case of a near-infrared image (for example, a wavelength region of about 800 nm to about 2500 nm), absorption at a specific wavelength in the near-infrared region can be imaged and analyzed using the distribution. In the case of a near-infrared image, it is possible to analyze moisture and oil content by wavelength. Further, in the case of a near-infrared image, absorption in the near-infrared region can be imaged as luminance information.

  Moreover, in the case of an infrared image (for example, a wavelength region is about 2500 nm-1000 micrometers), since a thermal image can be acquired, the temperature distribution state of the skin surface can be analyzed. Moreover, in the case of an infrared image, since a temperature change can be acquired, for example, a temperature change after massaging the skin, an influence on the external environment, and the like can be analyzed. Furthermore, in the case of an infrared image, the acquired image is obtained by converting temperature into luminance information. Therefore, the analysis for each mesh described above can be performed from the luminance information, and the above-described analysis method can be applied.

  Furthermore, in the case of an ultraviolet image (for example, a wavelength region of about 10 nm to about 400 nm), an image in which melanin is mainly emphasized can be acquired. Therefore, a spot portion is extracted and the skin is analyzed from the distribution state. be able to. That is, in the case of an ultraviolet image, an analysis specialized for spots can be performed. Furthermore, in the case of an ultraviolet image, since the acquired image is luminance information, the analysis for each mesh described above can be performed from the luminance information, and the analysis method described above can be applied.

  The above-mentioned visible image can be taken with a digital camera, for example, the near-infrared image can be taken with an imaging means for visualizing water or the like, the infrared image can be taken with, for example, a thermography, and the ultraviolet image is For example, the image can be taken with a camera or the like that emphasizes spots or the like, but is not limited thereto. In a broad sense, the technique in the first embodiment is an image obtained by quantifying a reflection or absorption state when an observation object is irradiated with electromagnetic waves. In addition, the amount of information changes depending on the size and number of blocks to be imaged. As for the illumination device, the photographing device, and the like, for example, a device shown in Japanese Patent No. 4882019 filed by the present applicant can be used, but the invention is not limited thereto.

  In addition, about each image mentioned above, a test subject's skin is analyzed using the test subject's face image image | photographed by the predetermined | prescribed wavelength range according to the analysis content etc.

<Difference in mesh shape>
Here, the difference in mesh shape for each subject in the first embodiment will be described with reference to the drawings. FIG. 4 is a diagram for explaining the difference in mesh shape. FIG. 4 includes face images 40-1 and 40-2 of different subjects as an example. Each of the face images 40-1 and 40-2 shows a front image of the face as an example. In the face area, a mesh group is arranged based on the reference points described above, and each area is divided by the mesh.

  Here, when the same mesh portion of each of the face images 40-1 and 40-2 is extracted as shown in FIG. 4, the size of the mesh of each of the partial images 41-1 and 41-2 is different. Specifically, each mesh 42-1 arranged in the face image 40-1 is 28 pixels in height × 25 pixels in width, whereas each mesh 42- arranged in the face image 40-2. 2 is 25 pixels vertical by 27 pixels horizontal. Since the mesh size is set and arranged according to the shape and size of the face for each subject as described above, the number of meshes for each subject becomes equal, and comparison between the meshes at the same position is performed with high accuracy. be able to. In the first embodiment, although the size of the mesh for each subject is different, since the comparison is performed based on statistics such as an average value and a standard deviation for each mesh, the influence as an evaluation index due to the difference in size is Absent.

<About input to the map>
Here, an example of inputting statistics to the map diagram in the first embodiment will be described. FIGS. 5 and 6 are diagrams (No. 1 and No. 2) for explaining an example of inputting statistics to the map diagram in the first embodiment. The example of FIG. 5 shows a front view of the face area, and the example of FIG. 6 shows left and right side views of the face area.

  In the first embodiment, for example, as shown in FIG. 5B, for each of the rectangular meshes 42 arranged in a predetermined number of lengths and widths with respect to the preset face area shown in FIG. The average brightness value and the standard brightness deviation value actually assigned by the subject are input. In the example of FIG. 5B, the average luminance value corresponding to the position of each mesh is input. The arrangement position and number of meshes are the same for each subject.

  When the statistics for the left and right face regions are measured, the statistics are input to the left and right meshes as shown in FIG. The statistics may not be input for all meshes, and may be input for some meshes.

  5 and 6, the model face image 43 as a sample is displayed instead of the face image of the subject itself, but the present invention is not limited to this, and the face image of the subject may be used. . In the first embodiment, since the number of meshes is the same for each subject, it is possible to input statistics at the same position even using a model face image 43 as shown in FIGS. Therefore, the analysis and evaluation integrated between each test subject can be performed. Thereby, in 1st Embodiment, analysis, such as a detailed moisture content and the dispersion | variation degree of the brightness | luminance between meshes, can be performed, for example per mesh.

  In the example of FIGS. 5 and 6, position information (for example, mesh number) is allocated to each mesh, whereby management for each mesh can be easily performed. Further, mesh information (for example, statistics) at the same position of a plurality of subjects can be easily extracted using the position information, and can be used for analysis and evaluation.

<Content of analysis>
Next, analysis contents from the statistics in the first embodiment will be described with reference to the drawings. FIG. 7 is a diagram illustrating an example of analysis content corresponding to a statistic. 7A shows an example of a moisture map based on the luminance average value, and FIG. 7B shows a skin variation map based on the luminance standard deviation.

  As shown in FIG. 7A, the average brightness value in the mesh is calculated by the above-described calculation of the statistic, so that the brighter one has less water content and the darker one has more water content. For this reason, in the first embodiment, “water content is high or low” is set based on a predetermined level according to the brightness level of the average brightness value, for example, a color map for each level level as a moisture map. Can be displayed.

  Also, with respect to the standard deviation of luminance, for example, when there is rough skin (shading), the luminance standard deviation becomes non-uniform. Therefore, in the first embodiment, “skin uniformity, non-uniformity” is set based on a predetermined level in association with the result of the standard deviation of luminance in the mesh, and for example, color-coded for each level level as a skin variation map. Can be displayed on the screen. The contents that can be analyzed using the luminance average value and the luminance standard deviation are not limited to this, and depending on the acquired image, for example, analysis and evaluation of skin smoothness, roughness, spots, etc. Etc. may be performed.

  The standard deviation value may be, for example, a standard deviation based on data of a plurality of subjects, or a standard deviation based on data statistically acquired from individual subjects. For example, in the case of using data statistically acquired from individual subjects, it is possible to perform skin analysis corresponding to specific skin characteristics, and to realize more accurate image analysis.

<Image generation example>
Next, an image generation example based on the analysis result will be described with reference to the drawings. FIG. 8 is a diagram illustrating an example of a moisture map. FIG. 9 is a diagram illustrating an example of a variation map. In the examples of FIGS. 8 and 9, mapping is performed based on a numerical table of face images to which statistics are input.

  For example, in the first embodiment, as shown in FIG. 8, a predetermined level (for example, Levels 1 to 5) is defined for the average value of luminance, and mapping is performed by color coding or the like based on the level. Specifically, in the example of FIG. 8, the levels are set as “Level 1: very high”, “Level 2: high”, “Level 3: slightly low”, “Level 4: low”, and “Level 5: very low”. ing.

  FIG. 8 shows statistical reference amounts (average luminance value, percentile value (%)) for each level. For example, in the example of FIG. 8, the luminance average value 0 to 68 is set as Level 1 with the percentile value of 20% as a delimiter. Similarly, the luminance average values 68 to 74 are set as Level 2 with the percentile value of 40% as a delimiter. Similarly, each of Levels 3 to 5 is set with the average luminance value as shown in FIG. Note that the number of levels and the values for dividing the levels in the present embodiment are not limited to the example of FIG. 8 and can be arbitrarily set.

  In this embodiment, as shown in FIG. 9, a predetermined level (for example, Levels 1 to 5) is defined for the standard deviation of luminance, and mapping is performed by color coding or the like based on the level. Specifically, in the example of FIG. 9, “Level 1: very uniform”, “Level 2: uniform”, “Level 3: slightly varied”, “Level 4: varied”, “Level 5: very varied. "Is set as a standard.

  Further, FIG. 9 shows the statistical reference amount (standard deviation value of brightness, percentile value (%)) for each level. For example, in the example of FIG. 9, the standard deviation values 0 to 2.9 of the luminance are set as Level 1 with the percentile value of 20% as a delimiter. Similarly, the luminance standard deviation values 2.9 to 3.6 are set as Level 2 with the percentile value of 40% as a delimiter. Similarly, in Levels 3 to 5, each level is set with a luminance standard deviation value as shown in FIG. 9 as a delimiter. Note that the number of levels and the values for dividing the levels in the present embodiment are not limited to the example of FIG. 9 and can be arbitrarily set.

  Note that the mapping is preferably color-coded for each level as described above. Further, for example, in the examples of FIGS. 8 and 9, the level calculation method is classified into five levels using the percentile values in an arbitrary group and is color-coded, but the number and colors to be classified are limited to this. It is not something. 8 and 9 show the level distribution for both the front face and the left and right faces. However, the present invention is not limited to this, and only the front face, only one of the left and right faces, or a part thereof is shown. May be a distribution map of only the face area.

  In addition, the skin condition can be analyzed from the pattern of the level distribution map as described above, for example, by recognition by pattern matching with a preset pattern. FIG. 10 is a diagram illustrating an example of a level distribution corresponding to the skin state.

  In the first embodiment, for example, “level distribution of people who feel the roughness” shown in FIG. 10 (A), “level distribution of people who feel a little bit”, shown in FIG. 10 (C), Compared with the level distribution chart obtained from the subject by setting in advance the “level distribution of people who do not feel too much”, the “level distribution of people who do not feel the roughness” shown in FIG. (Pattern matching) to analyze the skin condition of the subject.

  In the first embodiment, “N skin (healthy skin)” shown in FIG. 10 (E), “D skin (dry skin)” shown in FIG. 10 (F), and “O skin” shown in FIG. 10 (G). (Old oil skin / oily skin) ”,“ DO skin (mixed skin) ”shown in FIG. 10 (H), etc. are set in advance, and compared with the level distribution chart obtained from the subject, the subject's skin The state can be analyzed. The level distribution corresponding to the skin condition described above is stored in advance in the storage unit 13 or the like, for example, and is read out by the image analysis unit 18 described above and used for the analysis processing described above or the like. It is read by the generation means 19 and used for the image generation processing described above.

Second Embodiment: Image Analysis Device: Schematic Configuration Example
Next, a functional configuration example of the image analysis apparatus according to the second embodiment will be described with reference to the drawings. FIG. 11 is a diagram illustrating an example of a functional configuration of the image analysis apparatus according to the second embodiment. In the image analysis device 50 shown in FIG. 11, components having substantially the same functions as those included in the image analysis device 10 in the first embodiment described above are denoted by the same reference numerals. Detailed description is omitted.

  An image analysis apparatus 50 shown in FIG. 11 includes an input unit 11, an output unit 12, a storage unit 13, an image acquisition unit 14, a correction unit 15, an image enhancement unit 51, a mesh arrangement unit 16, and statistics. The calculation unit 17, the image analysis unit 18, the image generation unit 19, the transmission / reception unit 20, and the control unit 21 are included.

  In the second embodiment, the image enhancing means 51 is provided as compared with the first embodiment. The image enhancement unit 51 performs, for example, color correction, noise removal, edge enhancement, and darkness on the image corrected by the correction unit 15 (when the correction is not performed, the captured image acquired by the image acquisition unit 14). At least one emphasis process is performed among shadow emphasis for emphasizing a portion. Note that these enhanced images can be acquired using, for example, a preset filter or the like, but are not limited thereto.

  The mesh placement unit 16 performs mesh placement using the image obtained by the image enhancement unit 51, and the statistic calculation unit 17 calculates the statistic. Therefore, in the second embodiment, since the statistic for the enhanced image is calculated, a difference in statistic can be obtained as compared with the first embodiment.

  As described above, in the second embodiment, for example, when the moisture of the skin surface layer portion is visualized using an image photographed using a near-infrared camera or the like, and the skin condition of the subject is evaluated from the distribution state, the image is displayed. Emphasis can visualize the distribution more clearly.

  In the second embodiment, regarding the moisture of the skin, since the dark portion is absorbed by moisture, the luminance value equal to or lower than the average value of the subject is regarded as the “moisturized part”. As for drying, the degree of drying is also used as an index by dividing it into, for example, a predetermined number of levels (for example, 4 levels) so that the dusting state can be easily understood. Therefore, in the second embodiment, the same effect as that obtained by the first embodiment can be obtained, and “moisture” and “dry” can be obtained by setting a reference using the luminance value of the near-infrared image. Can be visualized.

<Image Analysis Device 50: Hardware Configuration>
Here, as the hardware configuration of the image analysis device 50 described above, for example, the same configuration as the hardware configuration of the image analysis device 10 described above can be used, and thus a specific description thereof is omitted here.

<Image Analysis Processing Procedure in Second Embodiment>
Next, an image analysis processing procedure in the second embodiment using the image analysis device 50 and the image analysis program described above will be described. FIG. 12 is a flowchart illustrating an example of an image analysis processing procedure according to the second embodiment.

  In the image analysis process shown in FIG. 12, first, a captured image is acquired (S11). The acquired image includes, for example, at least one of a visible region captured image, a near infrared region captured image, and an ultraviolet captured image, but is not limited thereto. Next, in the image analysis process, at least one correction process among color correction, contrast correction, brightness correction, and the like is performed on the image acquired in S11. For example, a plurality of images are analyzed under the same conditions. It is made possible (S12). In the second embodiment, the process of S12 need not be performed.

  Next, in the image analysis processing, a target emphasized image is acquired for the corrected image using a preset filter or the like (S13), and a reference point is set using the acquired image (S14). ). The subsequent S15 to S20 processes are the same as the processes in S04 to 09 of the image analysis processing procedure in the first embodiment described above, and a specific description thereof will be omitted here.

<Obtained image acquisition method>
Here, an example of the acquisition method of the emphasized image by the image enhancement means 51 mentioned above is demonstrated using figures. FIG. 13 is a diagram for explaining an example of a method for acquiring an emphasized image.

  As an enhanced image acquisition method, for example, an image (difference image) from which unevenness is removed by subtracting an image (moisture image) captured in a wavelength region 1950 nm from an image (base image) captured in a wavelength region 1300 nm. get.

  In addition, an image (moisture image) photographed in the wavelength region 1950 nm is subjected to noise removal and luminance gradient removal using a preset high-pass filter and low-pass filter, and a white portion is extracted from the average value of the analysis region. Acquire enhanced image. Here, the removal of the above-described luminance gradient will be described. Normally, light is emitted almost evenly by adjusting the light source, so it is regarded as a uniform state, but there may be a change that is different from the absorption of moisture such as shadows in a slightly uneven area. . Therefore, in order to correct the change as noise so that the brightness is corrected by image processing to be flat, that is, the brightness change as a result of uniform irradiation, the slope of the brightness is removed. . Specifically, when the target is inclined with respect to the light source, the correction is performed so that the difference between the portion far from the light source and the portion close to the light source is reduced. Moreover, as shown in FIG. 13, a difference image can extract a white part by the average value of an analysis area, and can acquire an emphasis image.

  In the second embodiment, by using the above-described enhanced image or the like, for example, a dark gray dot-like region can be analyzed as a dry portion, and a gray gradation portion can be analyzed as a moisture portion. be able to. In the second embodiment, analysis can be performed using one piece of information to be analyzed, or various types of analysis can be performed by combining a plurality of pieces of information.

<Analysis result in the second embodiment>
FIG. 14 is a diagram illustrating an example of an analysis result in the second embodiment. FIG. 14A shows an example of the analysis result of moisture unevenness, FIG. 14B shows an example of the analysis result of drying unevenness, and FIG. 14C shows the result of both moisture unevenness and drying unevenness. An example in which is displayed.

  In the example of FIG. 14A, for example, a portion where the amount of water is higher than the luminance average value of a predetermined selection area set in advance is expressed by a watermark such as a predetermined color (for example, blue) corresponding to the level. The predetermined selection area refers to, for example, a polygon selection area in which a part to be analyzed is arbitrarily selected by a polygon when performing analysis, and surrounds a part excluding eyes, nose, and mouth, for example. Is set. Further, in the example of FIG. 14B, a portion of uneven drying is determined in a predetermined color (corresponding to a level) with respect to an image obtained by filtering a predetermined selection region with a predetermined bandpass filter (CutOff (high frequency, low frequency)). For example, red). In the second embodiment, as an analysis result, for example, as shown in FIGS. 14A and 14B, the Poly region average value and the index 1 (absolute value integration / Poly) for moisture unevenness and dry unevenness. Area) and index 2 (accumulation of difference from average / Poly area) may be displayed. The Poly area means, for example, a polygon (for example, a predetermined polygon) area.

  Further, as shown in FIG. 14C, both moisture unevenness and dry unevenness can be expressed for the front face and the left and right faces. Thereby, each uneven part can be emphasized and displayed, and a user or a test subject can be made to recognize an uneven part easily.

  As described above, according to the present invention, skin and the like can be analyzed with high accuracy using a photographed image of a subject. In addition, according to this invention, skin evaluation can be carried out combining the map produced | generated by the analysis process mentioned above using the near-infrared image, the map by sensory evaluation etc. which were preset, for example. Thereby, services such as new skin care counseling and cosmetics provision for the subject can be provided.

  The preferred embodiments of the present invention have been described in detail above, but the present invention is not limited to such specific embodiments, and various modifications, within the scope of the gist of the present invention described in the claims, It can be changed.

DESCRIPTION OF SYMBOLS 10,50 Image analysis apparatus 11 Input means 12 Output means 13 Storage means 14 Image acquisition means 15 Correction means 16 Mesh arrangement means 17 Statistics calculation means 18 Image analysis means 19 Image generation means 20 Transmission / reception means 21 Control means 31 Input device 32 Output Device 33 Drive device 34 Auxiliary storage device 35 Memory 36 CPU
37 Network connection device 38 Storage means 40 Face image 41 Partial image 42 Mesh 43 Model face image

Claims (11)

In the image analysis method for analyzing the skin of the face of the subject using the face image of the subject photographed in a predetermined wavelength region,
An area setting step for setting the vertical and horizontal lengths of a mesh area in which a plurality of meshes are arranged based on at least three reference points including at least the corners of the eye obtained from the face image;
The vertical length of the mesh area is calculated by dividing the vertical width of the mesh area by the vertical number of meshes, and the horizontal width of the mesh area is calculated by dividing the horizontal width of the mesh area by the horizontal number of meshes. A mesh placement step of calculating a thickness and placing a plurality of polygonal meshes of the same type of the calculated size;
A statistic calculation step for calculating a statistic for each mesh obtained by the mesh placement step;
An image analysis step for analyzing the skin for each mesh based on the statistic obtained by the statistic calculation step;
And an image generation step of generating an image of an analysis result obtained by the image analysis step.   The image analysis method according to claim 1, wherein the face image includes a front face and at least one of left and right faces.   The image analysis method according to claim 1, wherein the predetermined wavelength region is a near infrared ray of about 800 nm to about 2500 nm. The mesh placement step includes:
The image analysis method according to any one of claims 1 to 3, wherein the shape of the mesh is set so that the number of meshes is the same with another subject based on the reference point. . The statistic calculation step includes:
5. The image analysis method according to claim 1, further comprising: calculating a statistic based on an average value of luminance or standard deviation of luminance of pixels included in each mesh. The image generation step includes
6. The image analysis method according to claim 1, wherein a map color-coded or pattern-coded based on a predetermined criterion is generated based on the statistic calculated in the statistic calculating step. .   The image analysis method according to claim 1, further comprising an image enhancement step of performing enhancement processing on an image of a subject photographed in the predetermined wavelength region. In the image analysis device for analyzing the skin of the face of the subject using the face image of the subject photographed in a predetermined wavelength region,
Based on three or more reference points including at least the corners of the eye obtained from the image, the vertical and horizontal lengths of the mesh region in which a plurality of meshes are arranged are set, and the vertical width of the mesh region is the number of vertical meshes. By dividing, the vertical length of the mesh is calculated, and the horizontal width of the mesh area is divided by the number of horizontal sides of the mesh to calculate the horizontal length of the mesh, and the same type of the calculated size Mesh arranging means for arranging a plurality of polygonal meshes of
Statistic calculation means for calculating a statistic for each mesh obtained by the mesh placement means;
Image analysis means for analyzing the skin for each mesh based on the statistics obtained by the statistics calculation means;
An image analysis apparatus comprising: an image generation unit configured to generate an image of an analysis result obtained by the image analysis unit.   The image analysis apparatus according to claim 8, wherein the face image includes a front face and at least one of the left and right faces.   The image analysis apparatus according to claim 8 or 9, wherein the predetermined wavelength region is a near infrared ray of about 800 nm to about 2500 nm.   The image analysis program for functioning a computer as each means which the image analysis apparatus as described in any one of Claims 8 thru | or 10 has.