KR102328759B1 - Method for Digitizing of Complexion-Describing Adjectives - Google Patents

Abstract

The present invention relates to a method for objectively quantifying adjectives expressing skin color and a method for evaluating changes in skin condition using the same. When using the method of the present invention, it is possible to objectively quantify the adjectives for expressing the skin condition, so that it can be used as an objective and reliable evaluation index, and it is possible to provide various and reliable methods for evaluating changes in skin condition. .

Description

Method for Digitizing of Complexion-Describing Adjectives

The present invention relates to a method for objectively quantifying adjectives expressing skin color and a method for evaluating changes in skin condition using the same.

The skin is the largest organ that covers the entire human body and is responsible for its function as an indicator to check the health status [1]. Skin color affects not only pigments such as melanin, hemoglobin, and bilirubin, but also the skin condition such as the thickness of the stratum corneum, secretion of oil or moisture. Therefore, in the evaluation of skin color, it is good to consider environmental factors such as genetics, gender, race, degree of UV exposure, seasonal change, and personal health condition [2]. Healthy skin not only can make a good impression on people, but also plays an important role in recognizing physical condition, attractiveness and age [1]. Women of various races tried to improve their skin condition, and this demand acted as a factor in developing the related cosmetic industry [3-5]. In particular, Asians have preferred white and white skin color compared to other races. This comes from the perception that bright, even-colored skin looks healthy, lively and youthful [6]. People use cosmetics or undergo other dermatological procedures to make their skin bright, bright and healthy [7]. There are many products to improve complexion, but it is necessary to check whether an accurate evaluation is being made on how effective these products are to improve complexion on our skin [8]. In previous studies, there were many cases of evaluating skin tone brightness, uniformity, and redness [9]. However, there have been no reports of studies evaluating the somewhat subjective feeling of skin color such as 'healthy', 'lively', or 'young looking'.

1. D. I. Perrett, D. M. Burt, I. S. Penton-Voak, K. J. Lee, D. A. Rowland, and R. Edwards, Symmetry and human facial attractiveness, Evol. Hum. Behav., 20:295 (1999). 2. I. D. Stephen, V. Coetzee, M. Law Smith, and D. I. Perrett, Skin blood perfusion and oxygenation color affect perceived human health. PLoS One, 4, e 50 (2009). 3. B. Fink and N. Neave, The biology of facial beauty, Int. J. Cosmet. Sci., 27;317 (2005). 4. B. Fink, K. Grammer, and R. Thornhill, Human (Homosapiens) facialattractiveness in relation toskin texture and color, Int. J. Comp. Psychol., Vol 115, 1:92 (2001). 5. R. Mulhern, G. Fieldman, T. Hussey, J.-L. Leveque, and P.pineau, Do cosmetics enhance female caucasiqn facial attractiveness, Int.J.Cosmet.Sci.,35;199(2003). 6. P. Burger, A. Landreau, S. Azoulay, T. Michel, and X. Fernandez, Skin whitening cosmetics: feedback and challenges in the development of natural skin lighteners, Cosmetics;3(4),36(2016). 7. B. Fink, K. Grammer, and P. J. Matts, Visible skin color distribution plays a role in the perception of age, attractiveness, and health in female faces, Evol.Hum.Behav.;27:433 (2006). 8. C. Musnier, P. Piquemal, P. Beau, and J. C. Pittet, Visual evaluation in vivo of ‘complexion radiance’ using the C.L.B.TTMsensorymethodology,Skin.Res.Technol.;10:50(2004). 9. A. Petitjean, J.-M. Sainthillier, S. Mac-Mary, P. Muret, B. Closs, T. Gharbi, and P. Humbert, Skin rediance: how to quantify validation of an optical method, Skin.Res.Technol.;13:2 (2007) . 10. K. Miyamoto, H. Takiwaki, S. Arase, and G. Hillebrand, Bioengineering of the skin, eds. K.-P. Wilhelm, E. Berardesca, P. Elsner, and H.I. Maibach, 209, Informa healthcare, New York (2006). 11. H. Takiwaki, H. Miyamoto, and K. Ahsan, A simple method to estimate CIE-L*a*b* values of the skin from its videomicroscopic image, Skin.Res.Technol.;3:42 (1997) .

The present inventors made intensive research efforts to develop an objective quantification method of adjectives expressing skin color in order to provide a more accurate and reliable method for evaluating complexion. As a result, when the quasi L*a*b* value, which is one of the indicators for expressing color, is used, it was found that adjectives expressing skin color can be quantified objectively and statistically significantly, thereby completing the present invention.

Accordingly, it is an object of the present invention to provide a method for objectively quantifying an adjective for skin condition expression including a predetermined step.

Another object of the present invention is to provide a method for evaluating changes in skin condition including predetermined steps.

According to one aspect of the present invention, there is provided a method for objectively quantifying an adjective for expressing a skin condition, comprising the following steps:

(a) preparing a random skin image;

(b) presenting an adjective for expressing a predetermined skin condition to an arbitrary panel, and adjusting the color of the skin image according to the subjective feeling conveyed by the adjective;

(c) extracting an average RGB value of the skin region of the color-adjusted skin image;

(d) selecting the obtained quasi L value as a reference value of the adjective by substituting the extracted average RGB value into Equation 1 below; However, the number of panelists in step (b) may be two or more, and when the number of panels is two or more, the average value of the obtained two or more quasi L values is selected as the reference value of the adjective:

[Formula 1]

quasi L = 23.2753298 + 0.09189237×R + 0.34734768×G - 0.0018069×G ² + 0.00000464×G ³ .

The present inventors made intensive research efforts to develop an objective quantification method of adjectives expressing skin color in order to provide a more accurate and reliable method for evaluating complexion. As a result, it was confirmed that when using the quasi L*a*b* value, which is one of the indicators for expressing color, it was possible to quantify the adjectives expressing skin color objectively and statistically significantly.

As used herein, the term 'adjective for skin condition expression' means any word that can be generally used to express and convey a subjective feeling about skin color, and specifically, for example, 'pale', 'clear' )', 'radiant', 'lively', 'healthy', 'rosy' and 'dull' does not

Each step of the present invention will be described in detail.

(a) preparing a random skin image

As the skin image used in the implementation of this step, the skin image of any panel selected for the implementation of step (b) or the skin image of another person can be used without limitation. The area of the skin region included in the total image area is not particularly limited, but it is preferable that the ratio of immunity occupied by the skin region is higher.

(b) presenting an adjective for expressing a predetermined skin condition to an arbitrary panel, and adjusting the skin image color according to the subjective feeling conveyed by the adjective

Any panel means a panel of one or two or more persons, which is not particularly limited. Adjustment of skin image color can be performed using, for example, various image editing programs, and specifically, for example, various versions of Photoshop (adobe company) programs can be used to adjust brightness, color, etc. It can be carried out without limitation by adjusting individually or by adjusting the curve value.

(c) extracting the average RGB value of the skin region of the color-adjusted skin image

The skin region of the skin image in this step means a region selected to include only the skin region among the entire image area, does not necessarily mean the entire skin region, and may be a partial region selected within the skin region. Specifically, for example, in one embodiment of the present invention, the skin area at this stage was defined as an “Area Of Interest (AOI)”, and as shown in FIG. 2, 6 cm x under the eyes An average RGB value was extracted by selecting a 6 cm area. The above-mentioned 6 cm x 6 cm area can be freely deformed as needed. The extraction method of the average RGB value is not particularly limited, but may be extracted using a program such as Image-Pro Plus (Ver.4.5.0.29, Media Cybernetics Inc., USA).

As used herein, the term “RGB” refers to the three primary colors of light, red (Red), green (Green), and blue (Blue), and each value may be represented in the range of 0 to 255. For example, R A color with a value of 255, a G value of 0, and a B value of 0 is red; a color with an R value of 0, a G value of 255, and a B value of 0 is green; a color with an R value of 0, a G value of 0, B A color with a value of 255 becomes blue.

(d) selecting the obtained quasi L value as a reference value of the adjective by substituting the extracted average RGB value into Equation 1 below; However, the number of panelists in step (b) may be two or more, and when the number of panels is two or more, the average value of the two or more quasi L values obtained is selected as the reference value of the adjective

[Formula 1]

quasi L = 23.2753298 + 0.09189237×R + 0.34734768×G - 0.0018069×G ² + 0.00000464×G ³ .

The present inventors have confirmed that, when the quasi L value obtained by the above-described formula 1 is used, a quasi L value without statistically significant deviation is obtained regardless of panel selection or image selection.

Therefore, the quasi L value obtained by the method of the present invention can be used as a numerical value that can objectively express adjectives for expressing various skin conditions.

According to another aspect of the present invention, there is provided a method for evaluating changes in skin condition comprising the following steps:

(a) a first skin image acquisition step of imaging a predetermined skin region of a subject requiring skin condition change evaluation;

(b) calculating a reference value for an adjective for expressing a predetermined skin condition by using the first skin image;

(b-1) presenting an adjective for expressing a predetermined skin condition to the subject, and adjusting the color of the first skin image according to the subjective feeling conveyed by the adjective;

(b-2) extracting an average RGB value of a skin region of the first skin image whose color has been adjusted;

(b-3) By substituting the average RGB value extracted in step (b-2) into the following formulas 1 to 3, at least one of the obtained quasi L, quasi a, and quasi b values is selected as the reference value of the predetermined adjective to do;

[Formula 1]

quasi L = 23.2753298 + 0.09189237×R + 0.34734768×G - 0.0018069×G ² + 0.00000464×G ³ ;

[Formula 2]

quasi a = -4.6472575 + 0.40731244×R - 0.4905761×G + 0.31679128×B - 0.0014044×R×B - 0.0010724×G×B + 0.00000657×RGB

[Formula 3]

quasi b = 0.3708721 + 0.18277815×R + 0.23190101×G 0.4344087×B

(c) a second skin image acquisition step of imaging the same skin region of the same subject at an arbitrary time interval from the first skin image acquisition step;

(d) extracting an average RGB value of the skin region of the first skin image obtained in step (a) and the second skin image obtained in step (c);

(e) Substituting the average RGB value of the skin region of the first skin image and the average RGB value of the skin region of the second skin image extracted in step (d) into Equations 1 to 3, each quasi L, quasi a and obtaining one or more of the quasi b values;

(f) at least one of quasi L, quasi a and quasi b values of the first skin image and at least one of quasi L, quasi a and quasi b values of the second skin image with respect to the reference value obtained in step (b) By comparing the values, respectively, the reference value and the quasi L, quasi a and quasi b values of the reference value and the second skin image based on the deviation of at least one of the reference value and the quasi L, quasi a, and quasi b values of the first skin image determining whether the deviation of at least one of the values decreases, and if the deviation decreases, evaluating that the skin condition has changed to a state closer to the predetermined adjective.

Hereinafter, the present invention will be described in detail step by step.

(a) a first skin image acquisition step of imaging a predetermined skin region of a subject requiring skin condition change evaluation

As used herein, the term “first skin image” refers to an image serving as a standard for evaluating whether or not the skin condition has changed. The area of the skin region included in the total image area is not particularly limited, but it is preferable that the ratio of immunity occupied by the skin region is higher.

(b) calculating a reference value for an adjective for expressing a predetermined skin condition by using the first skin image

This step includes the following substeps (b-1) to (b-3). Since this step uses the first skin image obtained in step (a), it should be carried out after step (a), but it is not necessarily carried out before step (c), after step (a) and It is understood that the implementation is possible at any time before the implementation of step (f).

(b-1) presenting an adjective for expressing a predetermined skin condition to the subject, and adjusting the color of the first skin image according to the subjective feeling conveyed by the adjective

As already described above in another aspect of the present invention, the skin image color adjustment can be performed using, for example, various image editing programs, and specifically, for example, various versions of Photoshop (adobe) programs are used. It can be carried out using the method, and it can be carried out without limitation by individually adjusting brightness, color, etc. or by adjusting the curve value.

(b-2) extracting the average RGB value of the skin region of the first skin image whose color has been adjusted

The definition of the term “skin region” and the method of extracting the average RGB value of the present invention are the same as those of the detailed description of the “objective quantification method of an adjective for skin condition expression”, which is an aspect of the present invention. , and the description thereof is omitted to avoid excessive complexity of the description of the present specification.

(b-3) By substituting the average RGB value extracted in step (b-2) into the following formulas 1 to 3, at least one of the obtained quasi L, quasi a, and quasi b values is selected as the reference value of the predetermined adjective step to do

[Formula 1]

quasi L = 23.2753298 + 0.09189237×R + 0.34734768×G - 0.0018069×G ² + 0.00000464×G ³

[Formula 2]

quasi a = -4.6472575 + 0.40731244×R - 0.4905761×G + 0.31679128×B - 0.0014044×R×B - 0.0010724×G×B + 0.00000657×RGB

[Formula 3]

quasi b = 0.3708721 + 0.18277815×R + 0.23190101×G 0.4344087×B

As used herein, the terms “quasi L”, “quasi a”, and “quasi b” refer to lightness, redness, and yellowishness, respectively. In particular, quasi a, which represents redness, indicates which side is biased toward red or green. If the quasi a value is negative, it indicates a color biased toward green, and if it is positive, it indicates a color biased toward red. quasi b indicates which side is biased between yellow and blue. If the quasi b value is negative, it means a blue-biased color, and if positive, it means a yellow-biased color.

Any one, two or all three values of quasi L, quasi a, and quasi b may be used as the reference value, and preferably, at least the quasi L value may be selected as the reference value, and quasi L as the reference value The value is used for comparison with the quasi L value of the first image and the second image, and the quasi a value serving as the reference value is used for comparison with the quasi a value of the first image and the second image, and the reference value is The quasi b value is used for comparison with the quasi L values of the first image and the second image.

(c) a second skin image acquisition step of imaging the same skin region of the same subject at an arbitrary time interval from the first skin image acquisition step

In the present invention, the term “arbitrary time interval” means a significant time interval for comparing whether and/or the degree of change in skin condition, and is natural and/or capable of causing a change in skin condition during that time. It may be accompanied by changes in anthropogenic conditions. Specifically, for example, changes in the above conditions include changes in the temperature or humidity of the surrounding environment, application of a prescribed cosmetic or external medicament to the skin, and administration of drugs for oral administration that can change the skin condition. can, but is not limited thereto. Even the above-mentioned “arbitrary time interval” may simply be a predetermined time interval for measuring a change in skin condition over time. The above-mentioned “any time interval” may be specifically, for example, minutes, hours, days, weeks, or months, or even years, as desired, and there is no particular limitation.

(d) extracting the average RGB value of the skin region of the first skin image obtained in step (a) and the second skin image obtained in step (c)

The definition of the term “skin region” and the method of extracting the average RGB value of the present invention are the same as those of the detailed description of the “objective quantification method of an adjective for skin condition expression”, which is an aspect of the present invention. , and the description thereof is omitted to avoid excessive complexity of the description of the present specification.

(e) Substituting the average RGB value of the skin region of the first skin image and the average RGB value of the skin region of the second skin image extracted in step (d) into Equations 1 to 3, each quasi L, quasi a and Steps to get one or more of the quasi b values

It has been described above that any one, two, or all three values of quasi L, quasi a, and quasi b may be used as reference values. At least one of quasi L, quasi a, and quasi b corresponding to the selected reference value is obtained through Equations 1 to 3.

(f) at least one of quasi L, quasi a and quasi b values of the first skin image and at least one of quasi L, quasi a and quasi b values of the second skin image with respect to the reference value obtained in step (b) By comparing the values, respectively, the reference value and the quasi L, quasi a and quasi b values of the reference value and the second skin image based on the deviation of at least one of the reference value and the quasi L, quasi a, and quasi b values of the first skin image determining whether the deviation of at least one of the values decreases, and if the deviation decreases, evaluating that the skin condition has changed to a state closer to the predetermined adjective

In the present invention, the term “deviation” refers to an absolute value of a difference between a reference value and a quasi L, quasi a, or quasi b value of the first or second skin image corresponding thereto. The smaller the deviation, the closer to the state meant by the corresponding adjective. Therefore, compared to the deviation between the quasi L, quasi a, or quasi b values of the first skin image and the reference value, the quasi L, quasi a of the second skin image Alternatively, when the deviation between the quasi b value and the reference value is further reduced, the skin of the subject may be evaluated as being changed to a state closer to the state indicated by the corresponding adjective.

In one embodiment of the present invention, the first skin image and the second skin image of the present invention are images that have undergone a color calibration process.

In the present specification, the term “calibration” refers to a process of adjusting quasi L, quasi a, and quasi b values for an output image to appear as a constant standard value. Specifically, for example, the first skin image and the second skin image may include a commercially available reference color panel for color calibration, for example, SpyderCHECKR (datacolor), etc. color calibration can be performed by The above-mentioned “output image” may be an image displayed on a monitor or printed on a two-dimensional plane with no material limitation, and there is no particular limitation.

The features and advantages of the present invention are summarized as follows:

(a) The present invention provides a method for objectively quantifying an adjective for skin condition expression, including a predetermined implementation step.

(b) The present invention provides a method for evaluating changes in skin condition including a predetermined implementation step.

(c) In the case of using the method of the present invention, it is possible to objectively quantify the adjectives for expressing the skin condition, which can be used as an objective and reliable evaluation index, and to provide a diverse and reliable method for evaluating changes in skin condition. make it possible

1 shows the standard images of a standard image (a) for dark skin (L* = 65) and a standard image for light skin (b) (L* = 74).
2 shows a schematic diagram of the process of manipulating the color of a standard image using the curve tool of Photoshop CS6. (a) shows the standard image (dark skin color), (b) shows the color-adjusted image, and (c) shows the analysis area.
3A to 3B show the grouping patterns of adjectives quantified as quasi L*, a*, and b* values (X-axis represents a manipulated value for a light skin image, and Y-axis represents a manipulated value for a dark skin image). 3A shows quasi L*, FIG. 3B shows quasi a*, and FIG. 3C shows quasi b* Probability p (ANOVA).

Hereinafter, the present invention will be described in more detail through examples. These examples are only for illustrating the present invention in more detail, and it will be apparent to those skilled in the art that the scope of the present invention is not limited by these examples according to the gist of the present invention. .

Example

Example 1: Selection of standard facial images

The image database of this research institute (Ellead Co., Ltd.) was searched to find the facial image, which is the standard for dark and light tones to be used in this study. As a result of analyzing the facial images of women, the quasi L ranged from 65 to 74, with the lowest image as a representative image of dark skin tone (quasi L=65) and the highest image as the lightest skin tone (quasi L=74). was selected as a representative image of In the selected representative image, only one part of the face was left so that it was impossible to identify the face, and when the skin color was adjusted according to the adjective (manipulation), the positions of the eyes and nose and the lines of the face were removed so as not to affect the appearance of the face. (See Fig. 1).

Example 2: CDA selection and image tone adjustment (Image manipulation)

In order to select CDAs (complexion-describing adjectives) to be used for evaluation, the following 7 adjectives were selected by examining words that are frequently exposed in various advertising media and are not difficult to use in daily life: 'pale', 'clear', 'radiant', 'lively', 'healthy', 'rosy', 'dull'. 30 researchers with experience in skin color evaluation at this research institute were selected as a panel, and the reference image of dark skin tone and light skin tone for each adjective was created using the Curve of Photoshop CS6 (Adobe system Incorporated, USA). was made to adjust. To reduce visual interference for each CDA, an interval of 90 to 120 seconds (s) was set between image tone adjustments.

Example 3: Image Analysis

The image analysis area (Area Of Interest; AOI) was fixed with an area of 6 cm x 6 cm under the eyes as shown in FIG. 2 . Using Image-Pro Plus (Ver.4.5.0.29, Media Cybernetics Inc., USA), the average RGB value of AOI was extracted from each CDA-adjusted image, and the quasi L, a, and b values were calculated using the following conversion formula. derived [10, 11].

Conversion of RGB values to quasi L*, a*, and b*:

quasi L* = 23.2753298 + 0.09189237×R + 0.34734768×G - 0.0018069×G2 + 0.00000464×G3

quasi a* = -4.6472575 + 0.40731244×R - 0.4905761×G + 0.31679128×B - 0.0014044×R×B - 0.0010724×G×B + 0.00000657×RGB

quasi b* = 0.3708721 + 0.18277815×R + 0.23190101×G 0.4344087×B

Example 4: Statistical Analysis

The paired t- test ( p<0.05 ) was used to verify the significance between the standard image and the adjusted image , and the independent t- test and the Mann-Whitney U test were used to verify the significance between the light-skin and dark-skin adjusted images. Significance between CDAs was confirmed through analysis of variance (ANOVA). Statistical processing was analyzed using SPSS version 21.0 (SPSS, USA).

Experiment result

The panel recognizes the CDA selected in advance, and draws the curve curve of Photoshop so that the skin color on the representative image of dark skin tone and representative image of light skin tone is corrected to a color suitable for the skin tone corresponding to the CDA that each panel thinks. was to be adjusted. As a result, in the case of representative images with dark skin tones, compared to standard images, images with CDA had quasi L values of 'pale', 'clean', 'bright', 'vigorous', 'healthy', and 'reddish'. There was a statistically significant increase, and on the contrary, the quasi L value of 'dark' decreased ( p <0.05). In the case of representative images with light skin tone, the quasi L values of 'pale', 'clean', 'bright', and 'vigorous' increased ( p<0.05 ), but the quasi of 'healthy', 'reddish' and 'dull' L values decreased ( p<0.05 ) (Table 1A).

In the case of quasi a and quasi b values, the CDA except for 'dull' in the representative image of darker skin tone than the standard image increased ( p<0.05 ), and in the representative image of light skin tone, 'vigorous', 'healthy', ' There was an increase in 'reddish' and 'dark' ( p<0.05 ) (Table 1B). In addition, in the case of quasi a and b, 'pale', 'clean', 'bright', 'vigorous', 'healthy', and 'reddish' in the representative images of dark skin tones, and 'pale' and 'clean' in light skin tones. , 'bright' was decreased compared to the values of quasi a and b of each representative image ( p<0.05 ) (Table 1C).

In addition, the difference between light skin tone and dark skin tone was checked to check whether the individual's perception of reflecting each CDA on the image differs depending on skin tone. As a result, it was confirmed that there was no statistically significant difference in all seven CDAs in the case of quasi L values indicating brightness.

In order to check whether the seven words selected for this study have a common relationship as a subjective expression of skin color, a significant difference ( p<0.05 ) was confirmed through statistical analysis between the image analysis values reflecting each CDA. As a result, it was confirmed that the groups were divided into (i) 'pale'-'clean'-'bright', (ii) 'vigorous'-'healthy'-'reddish', and (iii) dull (Table 2). and Fig. 3).

Claims (3)

delete A method for evaluating changes in skin condition comprising the following steps:
(a) a first skin image acquisition step of imaging a predetermined skin region of a subject requiring skin condition change evaluation;
(b) calculating a reference value for an adjective for expressing a predetermined skin condition by using the first skin image;
(b-1) presenting an adjective for expressing a predetermined skin condition to the subject, and adjusting the color of the first skin image according to the subjective feeling conveyed by the adjective;
(b-2) extracting an average RGB value of the skin region of the first skin image whose color has been adjusted;
(b-3) By substituting the average RGB value extracted in step (b-2) into the following formulas 1 to 3, at least one of the obtained quasi L, quasi a, and quasi b values is selected as the reference value of the predetermined adjective to do;
[Formula 1]
quasi L = 23.2753298 + 0.09189237×R + 0.34734768×G - 0.0018069×G ² + 0.00000464×G ³ ;
[Formula 2]
quasi a = -4.6472575 + 0.40731244×R - 0.4905761×G + 0.31679128×B - 0.0014044×R×B - 0.0010724×G×B + 0.00000657×RGB
[Formula 3]
quasi b = 0.3708721 + 0.18277815×R + 0.23190101×G 0.4344087×B
(c) a second skin image acquisition step of imaging the same skin region of the same subject at an arbitrary time interval from the first skin image acquisition step;
(d) extracting an average RGB value of the skin region of the first skin image obtained in step (a) and the second skin image obtained in step (c);
(e) Substituting the average RGB value of the skin region of the first skin image and the average RGB value of the skin region of the second skin image extracted in step (d) into Equations 1 to 3, each quasi L, quasi a and obtaining one or more of the quasi b values;
(f) at least one of quasi L, quasi a and quasi b values of the first skin image and at least one of quasi L, quasi a and quasi b values of the second skin image with respect to the reference value obtained in step (b) By comparing the values, respectively, the reference value and the quasi L, quasi a and quasi b values of the reference value and the second skin image based on the deviation of at least one of the reference value and the quasi L, quasi a, and quasi b values of the first skin image determining whether the deviation of at least one of the values decreases, and if the deviation decreases, evaluating that the skin condition has changed to a state closer to the predetermined adjective.
The method of claim 2, wherein the first skin image and the second skin image are images that have undergone a color calibration process.

Images