JP2016051444A - Evaluation method of aesthetic impression of lip and support device for supporting evaluation - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a technique for objectively quantizing aesthetic impression of a lip based on an image.SOLUTION: An evaluation method of aesthetic impression of a lip performed by a computer includes the steps of: obtaining an image on which an objective lip is imaged (S11); identifying a gloss part on the objective lip in the image (S12); and calculating at least one of, for example, an average area, area dispersion, a number, and the ratio of an outer periphery to an area (S13) as an index value correlated with aesthetic impression of a lip regarding a plurality of identified gloss parts.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a technique for analyzing an image showing lips.

  Patent Document 1 below proposes a method of generating synthetic face images with different stereoscopic effects. In this proposed method, principal component analysis is performed on the colors of the face images of a plurality of persons that have been standardized, and eigenvectors that contribute to the stereoscopic effect of the face are identified from among the eigenvectors obtained by principal component analysis. The following Patent Document 2 proposes a method for detecting the contours of the upper and lower lips of a face image and applying a highlight process to express the gloss of the lips, thereby performing a makeup process on the face image. . In the following Patent Document 3, a change in color or intensity is given to surface face image data composed of a reflected light component reflected on the surface of the subject's face, and the changed face image data and internal face image data are synthesized. A makeup simulation technique has been proposed. In Patent Document 4 below, the gloss of a face image that presents a plurality of chromatic gloss face images obtained by synthesizing a chromatic color component as a surface reflected light component with an internally scattered light face image with different hues of the surface reflected light components. Color presentation methods have been proposed. In this presentation method, the glossiness of each chromatic gloss face image is aligned.

Japanese Patent No. 5035524 Japanese Patent Laid-Open No. 2004-30007 JP 2003-256812 A JP 2008-6224 A

Each of the above-described methods does not present an evaluation method specialized for lips.
The lips have a rounded shape that protrudes outwardly, and vertical wrinkles and textures are formed on the surface thereof. From such a shape, vertical wrinkles and textures become rough or peeled due to external influences such as drying and ultraviolet rays. Therefore, various cosmetics for lips are provided.

  It is known that the impression of the mouth affects the impression of the entire face, and the aesthetic impression of the mouth, especially the lips, is a very important matter for women. What is desired for the aesthetic impression of the lips depends on the fashion, but now, lips that give a fresh, plump and three-dimensional impression are desired. Therefore, products that can give such aesthetic impression to cosmetics for lips are also provided.

  The present invention has been made in view of such problems, and provides a technique for objectively quantifying the aesthetic impression of lips based on images.

  Each aspect of the present invention employs the following configurations in order to solve the above-described problems.

  A 1st aspect is related with the evaluation method of the aesthetic impression of the lips performed by computer. The evaluation method according to the first aspect obtains an image in which a target lip is captured, specifies a glossy portion on the target lip in the image, and an index correlated with an aesthetic impression of the lips regarding the plurality of specified glossy portions Including calculating a value.

  A 2nd aspect is related with the assistance apparatus which assists evaluation of the aesthetic impression of a lip. The support device according to the second aspect includes an image acquisition unit that acquires an image of the target lip, a specifying unit that specifies a gloss part on the target lip in the image, and a plurality of specified gloss parts. A calculation unit that calculates an index value correlated with an aesthetic impression; and an output processing unit that performs an output process of the calculated index value.

  As another aspect of the present invention, a program that causes a computer to execute the evaluation method according to the first aspect described above may be used, or a computer-readable storage medium that records such a program may be used. . This recording medium includes a non-transitory tangible medium.

  According to each aspect described above, the aesthetic impression of the lips can be objectively quantified based on the image.

It is a flowchart which shows the evaluation method which evaluates the aesthetic impression of the lips in 1st embodiment. It is a figure for demonstrating the ratio of the outer periphery regarding a gloss part, and an area. It is a figure which shows the example which outputs an index value with a two-dimensional graph. It is a figure which shows notionally the hardware structural example of the assistance apparatus in 1st embodiment. It is a figure which shows notionally the process structural example of the assistance apparatus in 1st embodiment. It is a figure which shows two images which copied one test subject's lip to which 2 types of different cosmetics were each applied. It is a figure which shows typically the relationship between the aesthetic impression of lips, the amount of gloss, and the gloss shape. It is a flowchart which shows the evaluation method which evaluates the aesthetic impression of the lips in 2nd embodiment. It is a figure which shows notionally the process structural example of the assistance apparatus in 2nd embodiment. It is a figure which shows notionally the imaging | photography environment at the time of seeing a to-be-photographed object from the side. It is a figure which shows notionally the imaging environment at the time of seeing a camera direction from a to-be-photographed object side. It is a figure which shows the evaluation method of the aesthetic impression of the lips in an Example. It is a graph which shows the relationship between a stereoscopic effect score and the index value extracted from the image. It is a table | surface which shows the correlation coefficient of the index value extracted from the stereoscopic score and the image. It is a figure which shows the example of a relationship between the two-dimensional graph with two axes | shafts of a number and an average area, and a three-dimensional effect. It is a figure which shows the example of a relationship between the two-dimensional graph which has two axes of an average area and area dispersion, and a three-dimensional effect. It is a figure which shows the example of the time-dependent change of the three-dimensional effect represented by the two-dimensional graph with two axes | shafts of a number and an average area.

  Embodiments of the present invention will be described below. In addition, each embodiment given below is an illustration, respectively, and this invention is not limited to the structure of each following embodiment.

[First embodiment]
〔Evaluation method〕
FIG. 1 is a flowchart showing an evaluation method for evaluating the aesthetic impression of lips in the first embodiment. The evaluation method in the first embodiment is executed by a computer such as a support device described later. The evaluation method in the first embodiment includes a step of acquiring a lip image (S11), a step of specifying a glossy portion (S12), a step of calculating an index value (S13), and a step of outputting an index value (S14). .

  The aesthetic impression of the lips evaluated in this embodiment means the overall beauty that a person who sees the lips receives from the lips. Beauty includes cuteness and sexyness. More specifically, the beauty of the lips includes a soft and soft feeling that is accompanied by a rounded shape protruding outwardly. Such a specific impression of the lips is also called a three-dimensional effect.

  In step (S11), the computer acquires an image (denoted as a lip image) in which the target lips are shown. The acquired image data is acquired, for example, as an image file in JPEG (Joint Photographic Experts Group) format, BMP (Bitmap image) format, TIFF (Tagged Image File Format) format, GIF (Graphic Interchange Format) format, or the like. . However, the data format of the image is not limited, and may be a format in which the brightness information of each pixel is listed. Further, the acquired image may be a color image or a gray scale image obtained by converting the color image to gray scale. The computer may acquire the image from the camera that captured the lip image, or may acquire the image from another computer or a portable recording medium.

  The acquired image is preferably a surface reflected light image of the lips. This is because this evaluation method is intended for evaluation of the aesthetic impression of the lips accompanying the gloss of the lips. The gloss of the lips in the image is obtained by imaging the light reflected from the surface of the lips. The surface reflected light image can be acquired using polarized light, for example, as follows. The surface reflected light image is acquired from each polarization image obtained by projecting S-polarized light on the target lips (the lips of the subject) and imaging S-polarized light and P-polarized light. The computer can acquire each polarized image from another device, another computer, a portable recording medium, or the like, and use them to generate a surface reflected light image as the lip image.

  Specifically, the computer projects a polarized image obtained by projecting S-polarized light and imaging S-polarized light (SS polarized image), and a polarized image obtained by projecting S-polarized light and captured P-polarized light (S- The difference from the (P-polarized image) is taken. The SS polarized image is an image having a strong surface reflected light component of the target lip and a small amount of internally reflected light component, and the SP polarized image is an image having a strong internal reflected light component of the target lip. Therefore, by obtaining the difference between the SS polarized image and the SP polarized image, a surface reflected light image in which noise other than the surface reflected light component is further suppressed is acquired. Further, the computer further uses a PS polarization image obtained by projecting P polarization and imaging S polarization, and a PP polarization image obtained by projecting P polarization and imaging P polarization, A reflected light image may be acquired. Thereby, the surface reflected light image which further suppressed noises other than the surface reflected light component can be acquired. However, when using a P-S polarized image and a P-P polarized image, there is a problem that the photographing time becomes long, the subject is easily shifted between images, and more photographing equipment is required. is there. Therefore, according to the comparison between such a problem and the effect of noise suppression, a method using only the combination of the SS polarization image and the SP polarization image as described above is desirable. However, the method for acquiring the surface reflected light image is not limited.

  The target lips in the acquired image may be the lips of the subject or the lips in a state where cosmetics are applied. However, in this evaluation method, since the aesthetic impression of the lips accompanying the gloss of the lips is an evaluation target, it is desirable that the target lips are lips with a cosmetic applied. Further, the object to be imaged in the acquired image is not particularly limited as long as it includes lips. The image may include a person's part other than the lips and the background, or only a part of the lips. However, it is desirable that the acquired image is an image having a predetermined size at the center of the lower lip. This is because it has been found by the present inventor that the central portion of the lower lip tends to cause a difference in the aesthetic impression of the lips. The predetermined size may be set in a range that does not protrude the lower lip. Further, the shape of the acquired image is not limited to a square.

  The computer can cut out an image area corresponding to a predetermined size at the center of the lower lip in the acquired image, and acquire the image area as the lip image. In this case, the computer can automatically recognize the lips from the image using a known image recognition method. The computer can also specify a lip image region by displaying a screen that allows the user to specify a lip image region and detecting a user operation on the screen.

  In step (S12), the computer specifies a glossy part on the target lip in the image. “Glossy part” means a pixel group that strongly expresses the glossy component on the target lip in the image. The “non-glossy part” means a pixel group other than the pixels corresponding to the glossy part on the target lip in the image. Hereinafter, the pixels included in the glossy part may be referred to as “shiny pixels”, and the pixels included in the non-glossy part may be referred to as “non-shiny pixels”. In other words, the “shiny portion” means a set of four or eight connected gloss pixels, that is, a “shiny portion surrounded by a non-shiny portion”. However, the “shiny part” condition may be set more strictly. For example, a lower limit of the number of pixels may be added to the “shiny part” condition.

  The computer may execute binarization processing on the image area of the lips in the image acquired in (S11) in order to distinguish between glossy pixels and non-glossy pixels. The binarization threshold value used in the binarization process may be fixedly determined in advance, or may be calculated by a known method such as a mode method or a p-tile method. The computer may perform binarization processing on the converted grayscale image after converting the lip image into a grayscale image. In this case, the computer treats black pixels as non-glossy pixels and white pixels as glossy pixels. However, the computer can also determine the glossy pixel and the non-gloss pixel using the binarization threshold without performing the binarization process.

  In step (S13), the computer calculates index values related to the plurality of gloss parts specified in (S12). That is, the index value is a value calculated using information on the gloss part, and is a value having a correlation with the aesthetic impression of the lips. Specifically, the computer calculates, as the index value, at least one of an average area, an area variance, the number, and a ratio between the outer circumference and the area regarding a plurality of gloss parts. The number indicates the number of gloss parts. The average area is a value obtained by dividing the total number of pixels (area) included in the gloss part by the number. The area variance indicates the variance of the area of the gloss part, and is the mean square of the deviation of each area. The ratio between the outer periphery and the area is a ratio between the area of the gloss portion and the number of non-shine pixels forming the outer periphery. The ratio between the outer circumference and the area may be a ratio of the total area of the gloss part and the total outer circumference of the gloss part, or may be an average of the ratios calculated for each gloss part.

  FIG. 2 is a diagram for explaining a ratio between an outer periphery and an area relating to the gloss portion. In FIG. 2, glossy pixels are indicated by a symbol LP, and non-shine pixels are indicated by a symbol MP. The area S2 of the gloss portion is the number (9) of gloss pixels. The outer periphery Y is the number (12) of non-shiny pixels surrounding the periphery of the shiny part. In the example of FIG. 2, the number of gloss pixels that form a gloss part and the number of non-gloss pixels that share a piece on the outer circumference side thereof is calculated as the outer circumference Y. Moreover, the outer periphery Y may be calculated as the number of non-shiny pixels surrounding the periphery of the shiny part with four connections. In this case, the outer circumference Y is calculated as 16. In the example of FIG. 2, a value obtained by dividing the area S2 by the outer periphery is calculated as the ratio between the outer periphery and the area.

The computer may use the circularity of the gloss part calculated using the ratio between the outer periphery and the area as the index value. The circularity is a value indicating the closeness to the circular shape, and the circularity of the circle indicates the maximum value. The circularity can be calculated as follows, for example.
Circularity = area of gloss part (S2) / area calculated from outer periphery Y (S1)
S1 is an area calculated assuming that the outer periphery Y is a circle, and can be calculated by the following equation using the outer periphery Y.
S1 = πr ² = π (Y / 2π) ² = Y ² / 4π
Therefore, the circularity can be calculated by the following formula using the area S2 of the gloss part and the outer periphery Y.
Circularity = S2 / S1 = 4πS2 / Y ²
For example, the average circularity calculated for each gloss part is used as the index value.

  In (S14), the computer outputs the index value calculated in (S13). The computer may cause the index value to be displayed on the display device, or may cause the index value to be printed on the printing device. The computer may store the index value in an electronic file and send the electronic file to a portable recording medium or another device. The output mode of the index value is not limited.

The computer may output the index value itself, or may process and output the index value. For example, the computer uses two or three types of index values in the average area, area variance, number, and the ratio between the outer circumference and the area as axes, and the index values calculated in (S13) are plotted. Output a 3D or 3D graph.
FIG. 3 is a diagram illustrating an example in which the index value is output as a two-dimensional graph. In FIG. 3, the average area is set on the vertical axis and the number of gloss parts is set on the horizontal axis, and plotted at positions corresponding to the calculated average area and number.

  The evaluation method in the first embodiment evaluates the aesthetic impression of the target lip in the image acquired in (S11) based on the index value output in (S14). The actual evaluation based on the index value can be performed as follows, for example. Prepare in advance a correspondence table (including graphs) that shows the relationship between each index value calculated by the above evaluation method and the aesthetic impression score obtained from each lip image for multiple lips deep. The evaluator or the subject can estimate the aesthetic impression score of the target lip by comparing the index value output in (S14) with the correspondence table. When all of the average area, area variance, number, and the ratio between the outer circumference and the area are calculated, the evaluator or the subject estimates the aesthetic impression score for each index value, and from each score, the aesthetic impression of the target lip is calculated. It can also be evaluated. There are various specific evaluation methods based on the index value. In the second embodiment described later, other evaluation methods are exemplified.

[Supporting equipment]
FIG. 4 is a diagram conceptually illustrating a hardware configuration example of the support apparatus 10 in the first embodiment. The support apparatus 10 in the first embodiment is a so-called computer, and includes a CPU (Central Processing Unit) 1, a memory 2, an input / output interface (I / F) 3, a communication unit 4, and the like. The memory 2 is a RAM (Random Access Memory), a ROM (Read Only Memory), a hard disk, a portable storage medium, or the like. The communication unit 4 communicates with other computers via a communication network, exchanges signals with other devices, and the like. A portable recording medium or the like can be connected to the communication unit 4.

  The input / output I / F 3 can be connected to the output device 5, the input device 6, and the like. The output device 5 is a display device such as an LCD (Liquid Crystal Display) or a CRT (Cathode Ray Tube) display, a printer, or the like. The input device 6 is a device that receives an input of a user operation such as a keyboard and a mouse. However, the support device 10 may include a touch panel in which the output device 5 and the input device 6 are integrated. The support device 10 may include hardware elements not shown in FIG. 4, and the hardware configuration of the support device 10 is not limited.

The support device 10 in the first embodiment supports the evaluation of the aesthetic impression of the lips.
FIG. 5 is a diagram conceptually illustrating a processing configuration example of the support device 10 in the first embodiment. The support apparatus 10 in the first embodiment includes an image acquisition unit 11, a specification unit 12, a calculation unit 13, an output processing unit 14, and the like. Each of these processing units is realized, for example, by executing a program stored in the memory 2 by the CPU 1. The program is installed from 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 3 or the communication unit 4 and is stored in the memory 3. It may be stored.

  The image acquisition unit 11 executes (S11) described above. When a camera is connected to the input / output I / F 3 or the communication unit 4, the image acquisition unit 11 can acquire a lip image from the camera. The image acquisition unit 11 can also acquire the lip image via a communication unit 4 from a portable recording medium, another computer, or the like. The image acquisition unit 11 may acquire a surface reflected light image from the outside, or acquire each polarization image such as an SS polarization image or an SP polarization image from the outside, and take a difference between these polarization images. Thus, the surface reflected light image can also be calculated. In addition, the image acquisition unit 11 cuts out an image region representing the lips or an image region corresponding to a predetermined size at the center of the lower lip from the acquired image, and uses the cut out image as a lip image as another processing unit. May be used.

The specifying unit 12 executes (S12) described above. The specifying unit 12 can also execute binarization processing.
The calculation unit 13 executes (S13) described above.

  The output processing unit 14 executes (S14) described above. The output processing unit 14 causes the output device 5 to output the index value. The output processing unit 14 can also send the index value to a portable recording medium or another computer via the communication unit 4. The output processing unit 14 is a two-dimensional or tertiary model in which the calculated index values are plotted with the average area, the area variance, the number, and the two or three index values in the ratio between the outer circumference and the area as axes. You can also output the original graph.

[Operation and Effect in First Embodiment]
As described above, in the first embodiment, the gloss part on the target lip in the image is specified, and an index value correlated with the aesthetic impression of the lip related to the specified gloss part is calculated. Due to the correlation, the calculated index value is obtained by quantifying the aesthetic impression of the lips. Therefore, the aesthetic impression of the target lip appearing in the image can be objectively evaluated using the index value.

  FIG. 6 is a diagram showing two images showing the lips of one subject to which two different types of cosmetics are respectively applied. Comparing two lip images showing the entire lips, it is clear that the aesthetic impression of the left lip is more fresh and soft than the right lip. Here, comparing the lower two images obtained by cutting out a part of the lower lip, it can be seen that the left image has more gloss and the wrinkles separated by wrinkles than the right image. On the other hand, it can be seen that the right image has few glosses and many glosses separated by wrinkles, as emphasized by lines.

  From such a comparison, the present inventor paid attention to the amount of gloss of the lips and the shape of the gloss in the image as elements that influence the aesthetic impression of the lips. And this inventor discovered the following relations in the gloss amount and gloss shape between the lip with a comparatively good aesthetic impression, and the lip with a comparatively bad aesthetic impression.

FIG. 7 is a diagram schematically showing the relationship between the aesthetic impression of the lips, the gloss amount, and the gloss shape. In FIG. 7, lips with a relatively good aesthetic impression are shown above, and lips with a relatively bad aesthetic impression are shown below. Further, on the left side, a schematic diagram of the lips showing the glossy part and the non-glossy part is shown, and on the right side, a surface reflected light image of a part of the lower lip of each lip is shown. The image on the right side is an image obtained by cutting out 300 × 150 pixels at the center of the lower lip. As shown in the schematic diagram on the left side, lips with relatively good aesthetic impressions and bad lips have the following characteristics.
The area of each of the lips = glossy parts with relatively good aesthetic impression is large. Each shape of the gloss part is distorted. The areas of the gloss parts are dispersed. There are few gloss parts.
The area of each of the lips = glossy parts where the aesthetic impression is relatively bad is small. Each shape of the gloss part is circular. The area of each glossy part is the same. There are many shiny parts.

  As described above, in the first embodiment, the gloss portion that determines the quality of such an aesthetic impression is specified in the image, and the characteristics (average area, number, variance of area, outer periphery and area) regarding the gloss portion are specified. (Shape distortion or circular)) is extracted as an index value. Therefore, according to the first embodiment, it is possible to objectively evaluate the aesthetic impression of the lips using this index value.

[Second Embodiment]
In the second embodiment, an aesthetic impression score is calculated as an evaluation result for the target lips in the acquired image. Hereinafter, the evaluation method for evaluating the aesthetic impression of the lips and the support device for supporting the evaluation according to the second embodiment will be described focusing on the contents different from the first embodiment. In the following description, the same contents as those in the first embodiment are omitted as appropriate.

〔Evaluation method〕
FIG. 8 is a flowchart showing an evaluation method for evaluating the aesthetic impression of the lips in the second embodiment. The evaluation method in the second embodiment is executed by a computer such as a support device. The evaluation method in the second embodiment includes steps (S81) to (S85). Step (S81), step (S82) and step (S83) are the same as step (S11), step (S12) and step (S13) shown in FIG.

  In step (S84), the computer calculates the aesthetic impression score of the target lip in the lip image acquired in (S81) by inputting the index value calculated in (S83) to the multiple regression equation. The multiple regression equation is a multiple regression analysis using two or more index values as explanatory variables and the aesthetic impression score of the lips in the image as an objective variable. It is obtained by applying to an index value group and an aesthetic impression score group corresponding to human lips. The aesthetic impression score is obtained by quantifying the degree of quality of the aesthetic impression of the lips felt from the images when a plurality of evaluators look at the images in which the lips in each state are reflected.

  In order to increase the accuracy of the aesthetic impression score, it is desirable that there are many types of index values used for the explanatory variable of the multiple regression equation. For example, four types of index values such as average area, area variance, number, and the ratio between the outer circumference and the area are used as explanatory variables.

  Moreover, it is desirable to use the lips of the subject having the target lips in the lip image acquired in (S81) as the lips in the plurality of states to be subjected to the multiple regression analysis. Further, it is desirable that the lips of the plurality of lips to be subjected to the multiple regression analysis are lips whose shape characteristics (such as the shape of the lips, wrinkles of the lips, and texture) are approximate to the lips of the subject. . The computer may execute such a multiple regression analysis and may hold a multiple regression equation obtained as a result in advance or may be obtained from another computer.

  In (S85), the computer outputs the aesthetic impression score calculated in (S84). The computer may cause the display device to display the aesthetic impression score, or may cause the printing device to print the score. Further, the computer may store the aesthetic impression score in an electronic file, and send the electronic file to a portable recording medium or another device. The output mode of the aesthetic impression score is not limited.

  The computer may output only the aesthetic impression score, or may output the index value calculated in (S83) in addition to the aesthetic impression score. The output mode of the index value is as described in the first embodiment.

[Supporting equipment]
FIG. 9 is a diagram conceptually illustrating a processing configuration example of the support device 10 in the second embodiment. The support apparatus 10 in the second embodiment further includes a score calculation unit 16 in addition to the configuration of the first embodiment. The score calculation unit 16 is realized in the same manner as other processing units.

  The score calculation unit 16 executes the above (S84). The score calculation unit 16 may execute the multiple regression analysis as described above, and may hold a multiple regression equation obtained as a result thereof in advance or may be obtained from another computer.

  The output processing unit 14 executes (S85) described above. The output processing unit 14 causes the output device 5 to output an aesthetic impression score. The output processing unit 14 can also send the aesthetic impression score to a portable recording medium or another computer via the communication unit 4. The output processing unit 14 may output the index value together with the aesthetic impression score as in the first embodiment.

[Operation and Effect in Second Embodiment]
As described above, in the second embodiment, a plurality of types of index values calculated regarding the gloss part are input to the multiple regression equation, and as a result, the aesthetic impression score of the target lip in the acquired image is calculated. The multiple regression equation is a regression equation in which a plurality of types of index values correlated with the aesthetic impression of the lips are used as explanatory variables, and the aesthetic impression score obtained by sensory evaluation for the lips in the image is used as the objective variable. Therefore, according to the second embodiment, an objective quantitative value of the aesthetic impression of the target lip can be acquired as the aesthetic impression score.

[Supplement]
As described above, the target lips in the acquired image may be the lips of the subject or the lips in a state where cosmetics are applied. Applying lip cosmetics to the lips and executing the evaluation method as described above on the image of the lips to check the aesthetic impression of the lips, how to apply the lip cosmetics and the lip cosmetics And the like can be evaluated.

  Examples will be given below to describe the above-described embodiment in more detail. The contents of the above-described embodiments are not limited to the following contents.

  The relationship between the aesthetic impression and the gloss amount and gloss shape extracted from the lip image was confirmed as follows. First, when one person applies 18 kinds of cosmetics to one subject's lips, a plurality of SS polarization images and a plurality of surface reflected light images showing a plurality of states of lips relating to one lip are obtained. Prepared. For the 18 types of cosmetics, 9 types of stick type cosmetics and 9 types of liquid type cosmetics were used. A plurality of SS polarized images are respectively presented to five expert evaluators, and each expert evaluator gives a score of 0 points to 10 points for the aesthetic impression of each lip reflected in each SS polarized image. It was done. Each expert evaluator is an expert on the beauty of the lips and scored on the basis of the degree of perception of a fresh and plump impression as a concrete aesthetic impression. The fresh and plump impression obtained from the lips is often described as “three-dimensional effect” in recent years. Therefore, the score scored by the expert evaluator is hereinafter referred to as a three-dimensional score. “Stereoscopic” is a specific example of “aesthetic impression”.

The plurality of surface reflected light images were taken under the conditions shown in FIGS.
FIG. 10 is a diagram conceptually illustrating a shooting environment when a subject is viewed from the side.
FIG. 11 is a diagram conceptually showing a shooting environment when the camera direction is viewed from the subject side.
A round light source LS and a camera CM are installed at a distance of 0.7 meters (m) from the front of the subject, and an S-S polarized image is picked up using only the polarization filter PF1 that transmits the S-polarized component, and the polarization filter PF1. And an S—P polarization image is picked up using the polarization filter PF2 that transmits the P polarization component. The surface reflected light image is acquired by taking the difference between the SS polarization image and the SP polarization image. Further, a black lasha paper RP for light reduction was installed below the round light source LS. In this way, a lip image that easily causes a difference in the aesthetic impression (stereoscopic effect) of the lips is shot by illuminating the subject from the same direction as the shooting direction by the round light source LS placed substantially the same as the camera CM. be able to.

Specific conditions are as follows.
Camera CM: Nikon D300S
Light source: 1 circle (circular fluorescent lamp) (1/3 reduction with black lasha paper RP)
Lens: Micro Nikkor AiAf60mm f2.8D
Magnification ratio: 1 to 10
Distance: 0.7m
Shutter speed: 1/4 (when photographing an S-polarized image), 1 / 1.3 (when photographing an SP-polarized image)
White balance (WB): PRE d-1
Output: JPEG format Sensitivity: ISO200
Room lighting: Off

  FIG. 12 is a diagram illustrating a method for evaluating the aesthetic impression of the lips in the example. In the present embodiment, the center part of the lower lip (hereinafter referred to as a lip image) is cut out from each surface reflected light image acquired as described above, and binarized with a threshold value “40” for this lip image. Processing was performed, and from each binarized image, the average area, area variance, number and circularity of the gloss part, and the total area of the gloss part were calculated. The total area of the gloss portion is the total number of white pixels in the lip image. Further, in this embodiment, the luminance of each lip image is averaged, and the luminance is acquired from each averaged image. The luminance was measured as an L value with a measuring device (such as a color difference meter). The other index value calculation methods are as described in the first embodiment.

FIG. 13 is a graph showing the relationship between the stereoscopic effect score and the index value extracted from the image. (A) is the relationship between the brightness and the stereoscopic score, (B) is the relationship between the total area of the gloss part and the stereoscopic score, (C) is the relationship between the average area of the gloss part and the stereoscopic score, (D). Is the relationship between the circularity of the gloss portion and the stereoscopic score, (E) is the relationship between the number of gloss portions and the stereoscopic score, and (F) is the relationship between the area dispersion of the gloss portion and the stereoscopic score.
FIG. 14 is a table showing the correlation coefficient between the stereoscopic effect score and the index value extracted from the image.
The circularity shown in FIGS. 13 and 14 is an average value of the circularity of the gloss part.

  As shown in FIG. 13 and FIG. 14, there is a high correlation between the average area, the number, the circularity, and the area variance of the gloss part and the stereoscopic score, and the total area and luminance of the gloss part correlate with the stereoscopic score. Was confirmed to be low. Further, it was confirmed that the average area and number were more correlated than the area dispersion and the circularity. It was also confirmed that there was a positive correlation between the average area and area variance and the stereoscopic score, and a negative correlation between the number and circularity and the stereoscopic score. According to the present Example, it was proved that the aesthetic impression (three-dimensional effect) of the lips can be appropriately evaluated by the average area, number, circularity, and area dispersion regarding the gloss part.

  FIG. 15 is a diagram illustrating an example of a relationship between a two-dimensional graph having two axes of number and average area and a stereoscopic effect. In FIG. 15, the number of gloss parts is assigned to the x-axis, and the average area of the gloss parts is assigned to the y-axis. A stereoscopic effect score is attached to the points plotted on the two-dimensional graph. FIG. 15 shows that the stereoscopic effect score is high in the upper left area of the two-dimensional graph and the stereoscopic effect score is low in the lower right area of the two-dimensional graph. By plotting the number and average area calculated for the target lip on this two-dimensional graph, the stereoscopic effect score of the target lip can be estimated with high accuracy.

  FIG. 16 is a diagram illustrating an example of a relationship between a two-dimensional graph having two axes of average area and area dispersion and a stereoscopic effect. In FIG. 16, the average area of the gloss part is assigned to the x-axis, and the area variance of the gloss part is assigned to the y-axis. A stereoscopic effect score is also attached to the points plotted on the two-dimensional graph. According to FIG. 16, the stereoscopic effect score is high in the upper right area of the two-dimensional graph, and the stereoscopic effect score is low in the lower left area of the two-dimensional graph. By plotting the average area and area variance calculated for the target lip on this two-dimensional graph, the three-dimensional score of the target lip can be estimated with high accuracy.

  In this way, by plotting the index value calculated for the target lip on a two-dimensional graph with two types of index values correlated with the stereoscopic score as an axis, the aesthetic impression (three-dimensional effect) of the target lip is highly accurate. Can be evaluated.

  FIG. 17 is a diagram illustrating an example of a change in stereoscopic effect with time represented by a two-dimensional graph having two axes, the number and the average area. FIG. 17 has the same x-axis and y-axis as FIG. The relationship between each region of the two-dimensional graph and the stereoscopic effect score is the same as in FIG. That is, the stereoscopic effect score is high in the upper left region, and the stereoscopic effect score is low in the lower right region of the two-dimensional graph. Therefore, as shown in FIG. 17, each index value calculated from each image obtained by imaging the target lip at different times is plotted on the two-dimensional graph, so that the stereoscopic effect score of the target lip changes over time. Can be guessed. FIG. 17 shows changes over time in the index values calculated for the lips of two people. According to the graph of FIG. 17, it is estimated that the lips plotted in a circle are less deteriorated with time in the stereoscopic effect than the lips plotted in a square.

  In addition, in the some flowchart used by the above-mentioned description, although several process (process) is described in order, the execution order of the process performed by each embodiment is not restrict | limited to the order of the description. In each embodiment, the order of the illustrated steps can be changed within a range that does not hinder the contents. Each process may include contents executed by a person. For example, the images acquired in (S11) and (S81) may be read by a computer from a file selected by a human operation on the computer. Further, when the center part of the lower lip is cut out from the acquired image, the position of the center part may be designated by a human operation on the computer. Moreover, each above-mentioned embodiment can be combined in the range in which the content does not conflict.

  Part or all of the above embodiments and examples can be specified as follows. However, each above-mentioned embodiment and each Example are not restrict | limited to the following description.

<1> In the evaluation method of the aesthetic impression of the lips executed by a computer,
Get an image of the target lips,
Identify the glossy part on the target lip in the image,
Calculating an index value correlated with the aesthetic impression of the lips with respect to the plurality of specified gloss portions;
Evaluation method including that.

<2> The calculation of the index value is performed by calculating, as the index value, at least one of an average area, an area variance, a number, and a ratio between an outer periphery and an area regarding the plurality of specified gloss parts.
The evaluation method as described in <1>.
<3> An average area, area variance, number, and two-dimensional or three-dimensional index values plotted with respect to the target lip, with two or three index values in the ratio between the outer circumference and the area as axes. Output a three-dimensional graph,
The evaluation method according to <2>, further comprising:
<4> Each index value calculated from each image obtained by imaging the target lips at different times is plotted on the graph.
The evaluation method according to <2> or <3>, further comprising:
<5> A multiple regression analysis using two or more kinds of index values related to the gloss part as explanatory variables and an aesthetic impression score of the lips in the image as a target variable, Using the multiple regression equation obtained by applying to the index value group and aesthetic impression score group corresponding to the lips of a plurality of people, to calculate the aesthetic impression score of the target lips,
The evaluation method according to any one of <1> to <4>, further including:
<6> Binarize the image.
Further including
The non-shiny part is a black pixel group, and the shiny part is a white pixel group.
The evaluation method according to any one of <1> to <5>.
<7> The acquired image is an image that displays a predetermined size of the center of the lower lip.
The evaluation method according to any one of <1> to <6>.
<8> The acquired image is a surface reflected light image of the target lips.
The evaluation method according to any one of <1> to <7>.
<9> In a support device that supports the evaluation of the aesthetic impression of the lips,
Image acquisition means for acquiring an image showing the target lips;
Identifying means for identifying a glossy part on the target lip in the image;
Calculating means for calculating an index value correlated with the aesthetic impression of the lips for the specified plurality of gloss parts;
Output processing means for performing output processing of the calculated index value;
A support device comprising:
<10> The calculation unit calculates, as the index value, at least one of an average area, an area variance, a number, and a ratio between an outer periphery and an area regarding the plurality of specified gloss parts.
The support device according to <9>.
<11> The output processing means includes an average area, an area variance, a number, and two or three index values in the ratio between the outer periphery and the area, and the index value calculated for the target lip is Output a plotted 2D or 3D graph,
<10> The support device according to item 10.
<12> The output processing means plots each index value calculated from each image obtained by imaging the target lips at different times on the graph.
<10> or the support apparatus as described in <11>.
<13> A multiple regression analysis in which two or more kinds of index values relating to the gloss part are used as explanatory variables and an aesthetic impression score of the lips in the image is used as an objective variable. Score calculating means for calculating an aesthetic impression score of the target lip using a multiple regression equation obtained by applying to an index value group corresponding to a plurality of lips and an aesthetic impression score group;
The support device according to any one of <9> to <12>, further including:
<14> The specifying unit binarizes the image, specifies a gloss part on the target lip in the binarized image,
The non-shiny part is a black pixel group, and the shiny part is a white pixel group.
The support device according to any one of <9> to <13>.
<15> The image acquisition means acquires an image displaying a predetermined size of the center part of the lower lip.
The support device according to any one of <9> to <14>.
<16> The image acquisition means acquires a surface reflected light image of the target lips.
The support device according to any one of <9> to <15>.
<17> A program that causes a computer to execute the evaluation method according to any one of <1> to <8>.
<18> A method of applying a lip cosmetic to the lips and confirming an aesthetic impression by the evaluation method according to any one of <1> to <8>.

1 CPU
2 Memory 3 Input / output I / F
4 Communication Unit 5 Output Device 6 Input Device 10 Support Device 11 Image Acquisition Unit 12 Identification Unit 13 Calculation Unit 14 Output Processing Unit 16 Score Calculation Unit

Claims (11)

In the method for evaluating the aesthetic impression of lips executed by a computer,
Get an image of the target lips,
Identify the glossy part on the target lip in the image,
Calculating an index value correlated with the aesthetic impression of the lips with respect to the plurality of specified gloss portions;
Evaluation method including that. The calculation of the index value is to calculate at least one of an average area, an area variance, a number, and a ratio between the outer periphery and the area as the index value for the plurality of specified gloss parts.
The evaluation method according to claim 1. The average area, area variance, number, and 2 or 3 index values in the ratio between the outer circumference and the area are used as axes, and the index values calculated for the target lips are plotted in two or three dimensions. Output a graph,
The evaluation method according to claim 2, further comprising: In the graph, each index value calculated from each image obtained by imaging the target lips at different times is plotted, respectively.
The evaluation method according to claim 2 or 3, further comprising: Multiple regression analysis using two or more kinds of index values related to the gloss as an explanatory variable and the aesthetic impression score of the lips in the image as an objective variable. Using the multiple regression equation obtained by applying the index value group corresponding to the lips and the aesthetic impression score group, the aesthetic impression score of the target lips is calculated.
The evaluation method according to any one of claims 1 to 4, further comprising: Binarize the image;
Further including
The non-shiny part is a black pixel group, and the shiny part is a white pixel group.
The evaluation method according to any one of claims 1 to 5. The acquired image is an image displaying a predetermined size of the center part of the lower lip.
The evaluation method according to any one of claims 1 to 6. The acquired image is a surface reflected light image of the target lip.
The evaluation method according to any one of claims 1 to 7. In a support device that supports the evaluation of the aesthetic impression of the lips,
Image acquisition means for acquiring an image showing the target lips;
Identifying means for identifying a glossy part on the target lip in the image;
Calculating means for calculating an index value correlated with the aesthetic impression of the lips for the specified plurality of gloss parts;
Output processing means for performing output processing of the calculated index value;
A support device comprising:   The program which makes a computer perform the evaluation method of any one of Claim 1 to 8.   A method for applying a lip cosmetic to the lips and confirming an aesthetic impression by the evaluation method according to claim 1.