KR101998211B1 - Method for generating skin glossiness index - Google Patents

Abstract

A method of generating a skin luster index comprises: photographing a cross-polarization image of the skin; Imaging a parallel-polarization image of the skin; Obtaining a surface reflected light image corresponding to a difference between a parallel polarized image and a crossed polarized image; Detecting at least one light object from the surface reflected light image; And calculating the glossiness index using the number and / or area of the optical objects of the predetermined size or more among the one or more optical objects. By the method of generating the skin luster index, the skin luster can be objectively quantified and evaluated, and the beauty information related to the skin luster can be provided.

Description

METHOD FOR GENERATING SKIN GLOSSINESS INDEX [0002]

Embodiments relate to a method of generating a skin gloss index.

Due to the recent trend of making light and transparent make-up trends and the increasing number of customers who feel uncomfortable with make-up making frequent due to the busy life cycle, there is a demand for suppression of shine from make- Is increasing.

The shininess, that is, the shine of the skin, has been mainly recognized by the influence of reflected light in the inside and outside of the face and the microstructure of the face. Conventionally, a method of evaluating luster by separating images taken by a skin on a frequency basis, or evaluating glossiness by the number of pixels and brightness of image data of an image has been used. However, the results of this method have limitations that are not in good agreement with the naked skin condition.

Published Patent Publication No. 2001-0114041

According to an aspect of the present invention, it is possible to provide a method of generating a skin gloss index capable of objectively numerically evaluating skin gloss.

According to an embodiment, a method of generating a skin luminous index comprises: photographing a cross-polarization image of skin; Imaging a parallel-polarization image of the skin; Obtaining a surface reflected light image corresponding to a difference between a parallel polarized image and a crossed polarized image; Detecting at least one optical object from a surface reflected light image; And calculating a glossiness index using the optical object of the predetermined size or more among the one or more optical objects.

The step of detecting one or more optical objects may comprise detecting a pixel having a brightness above a predetermined threshold in the surface reflected light image.

The step of calculating the glossiness index may include calculating a value obtained by dividing the number of optical objects having a predetermined size or more by the number of optical objects excluding the optical object having a predetermined size or more among the one or more optical objects.

The step of calculating the luster index may include calculating an area of the optical object larger than a predetermined size.

The step of calculating the glossiness index may include a step of giving a predetermined weight to each of the above-described values and areas and adding them. The adding and weighting step may include a step of determining a weight through a regression analysis with a visual evaluation result.

According to an aspect of the present invention, it is possible to objectively quantify and evaluate the shine of the skin, that is, the shine, and provide a method of generating a skin gloss index that is in good agreement with the shine condition of the skin evaluated by the naked eye. In addition, cosmetic information related to the possibility of shininess may be provided by using the glossiness index generated by the skin glossiness index generation method.

FIG. 1 is a flowchart of a method of generating a skin luminous intensity index according to an embodiment.
Figure 2a is a cross-polarization image showing skin changes before and after matt-make-up.
FIG. 2B is a cross-polarized image showing the skin changes before and after the luminous makeup.
3A is a parallel-polarization image showing skin changes before and after matte makeup.
FIG. 3B is a parallel polarized image showing skin changes before and after the luminous makeup.
4A is a surface reflected light image showing skin before makeup according to one embodiment.
4B is a surface reflected light image showing skin after matte makeup according to an embodiment.
4C is a surface reflected light image showing the skin after glossy makeup according to an embodiment.
5 is a graph showing the results of visual evaluation of skin gloss.
6A is a surface reflected light image generated according to an embodiment.
FIG. 6B is an image showing a light object extracted from the surface reflected light image of FIG. 6A. FIG.
FIG. 7 is a graph showing the skin gloss index calculated according to an embodiment in comparison with a visual evaluation result. FIG.
FIG. 8 is a graph showing the skin gloss index calculated according to another embodiment by comparing it with a visual evaluation result.
FIG. 9 is a graph showing the skin gloss index calculated according to another embodiment by comparing it with a visual evaluation result. FIG.

Hereinafter, embodiments will be described in detail with reference to the drawings. However, the present invention is not limited by the following examples.

FIG. 1 is a flowchart of a method of generating a skin luminous intensity index according to an embodiment.

Referring to FIG. 1, a cross-polarization image of the skin of a subject to be calculated first can be photographed (S1). The cross-polarized image refers to an image obtained by causing the light of the light source and the light detected by the image to have a polarization direction orthogonal to each other. For example, a polarizing plate may be disposed on the light source and the camera, respectively, and the polarizing direction of the polarizing plates may be arranged so as to have a difference of 90 degrees.

In addition, a parallel-polarization image of the subject's skin can be photographed (S2). A parallel polarized image refers to an image obtained by causing the light of the light source and the light detected by the image to have the same polarization direction to each other. For example, after a polarizing plate having the same polarization direction is disposed on each of the light source and the camera, light incident on the camera can be taken as a parallel polarized image. The parallel polarized image can be photographed at the same point on the skin of the same subject that has photographed the cross polarized image in step S1 described above.

The cross-polarized image and the parallel-polarized image may be obtained using various imaging equipment, and are not limited to specific equipment. Also, in the flowchart of FIG. 1, it is shown that the photographing (S2) of the parallel polarized light image is performed after the photographing (S1) of the crossed polarized image is performed, but the steps are arranged arbitrarily for convenience of explanation. That is, in other embodiments, the parallel polarized image may be photographed prior to the cross polarized image, or the imaging of the cross polarized image and the parallel polarized image may be substantially simultaneous.

In one embodiment, the cross-polarized image and the parallel-polarized image may be obtained for regions of the ball portion where the effect of the shadow due to the skin structure, such as nose or lips, in the image taken of the face is relatively small. However, this is an example, and in another embodiment, crossed polarized images and parallel polarized images may be obtained by targeting other portions of the face or the entire face. Alternatively, a cross-polarized image and a parallel-polarized image may be obtained with respect to the entire face, and a luster index may be calculated by performing a subsequent image processing process on only a specific portion of the face in the image.

FIG. 2A is a cross-polarized image showing skin changes before and after a make-up, and FIG. 2B is a cross-polarized image showing skin changes before and after a makeup. That is, FIG. 2A shows the crossed polarized image obtained by photographing the skin after the matte makeup, minus the crossed polarized image obtained by photographing the skin before the matte makeup. Similarly, FIG. 2B shows the crossed polarized image obtained by photographing the skin after the glossy makeup, minus the crossed polarized image obtained by photographing the skin before the glossy makeup. As a result, Figs. 2A and 2B correspond to conversion of only the characteristic changes of the pixels such as brightness or color before and after make-up in the image into an image.

In the present specification, matte makeup refers to application of a coating or cosmetic having no or relatively low gloss to the skin, and glossy makeup refers to application of a coating or cosmetic having a relatively high gloss to the skin. The matte makeup and the milky makeup can be distinguished based on various characteristics such as the type of constituent material, viscosity, and moisture content. The matte makeup and the milky makeup in the present specification are defined as relative concepts for distinguishing each other and are not intended to mean absolute composition or other characteristics of the material. In the present specification, the matte makeup and gloss makeup are used only for the purpose of providing reference data for calculating the glossiness index, and the method of generating the glossiness index according to the embodiments is limited by the kind, characteristics, It is not.

When observing with the naked eye, the luminous make-up has a feature that the surface of the skin is shiny and has the effect of making the surface of the skin rough. However, the cross-polarized image shown in Figs. 2A and 2B showing the change before and after makeup is composed of components due to the reflected light inside the facial structure rather than the component due to the reflected light of the surface, and thus the difference between the matte makeup and the milky makeup .

In one embodiment, the brightness of each pixel of the image can be quantified and compared after converting the cross-polarized image shown in Figs. 2A and 2B to a gray scale. The quantification process can be performed using any image processing software that is known or will be developed in the future, and is not limited to a specific method. For example, image processing used in the embodiments may be performed using a commercially available image processing software, such as Image Pro Plus (Image-Pro Plus ^®). As a result of the image processing, the average value of the pixel brightness in the cross polarization image shown in FIG. 2A was about 4.35 (arbitrary unit), and the standard deviation was about 5.38. Also, in the cross-polarized image shown in Fig. 2B, the average value of the pixel brightness was about 4.35, and the standard deviation was about 5.07.

On the other hand, FIG. 3A shows a parallel polarized light image showing skin changes before and after the matte makeup, and FIG. 3B shows a parallel polarized image showing skin changes before and after the luminous makeup. Unlike the cross polarized light image shown in Figs. 2A and 2B, the parallel polarized light image shown in Figs. 3A and 3B contains a large amount of components due to the reflected light of the skin surface relatively, and the change in the skin luster due to makeup is relatively large . It can be seen that the parallel polarized image of FIG. 3B showing the skin changes before and after the milky make-up contains a lot of bright pixels compared to the parallel polarized image of FIG. 3A showing skin changes before and after matte makeup.

In one embodiment, the parallel polarized light image shown in Figs. 3A and 3B is converted to gray scale by the same method as the above-described numerical conversion process with reference to Figs. 2A and 2B, and then the brightness of each pixel of the image is digitized and compared . As a result, the average value of the pixel brightness in the parallel parallel image shown in FIG. 3A was about 3.57 (arbitrary unit), and the standard deviation was about 8.75. Also, in the parallel polarized light image shown in FIG. 3B, the average value of the pixel brightness was about 5.21, and the standard deviation was about 14.2.

Comparing FIG. 2 and FIG. 3, skin changes before and after makeup appear prominently with increasing average and standard deviation of pixel brightness in a parallel polarized image. Therefore, referring again to FIG. 1, a surface reflected light image showing the shine of the skin can be obtained based on the difference between the parallel polarized image and the crossed polarized image in which the same skin is photographed (S3). The surface reflection light image may be an image obtained by subtracting a cross polarization image from a parallel polarization image. For example, after subtracting the brightness value of each pixel of the cross-polarized image converted to gray scale from the brightness value of each pixel of the parallel-polarized image converted to gray scale, each pixel having brightness obtained as a result of the subtraction is re- . ≪ / RTI >

4A is a surface reflected light image showing skin before makeup. FIG. 4B is a surface reflected light image showing the skin after the matte makeup, and FIG. 4C is a surface reflected light image showing the skin after the milky makeup. In the surface reflected light image shown in Fig. 4A, the average value of the pixel brightness is about 3.30, and the standard deviation is about 4.49. In addition, in the surface reflected light image shown in Fig. 4B, the average value of the pixel brightness is about 3.17, and the standard deviation is about 7.67. Further, in the surface reflected light image shown in FIG. 4C, the average value of the pixel brightness is about 4.93, and the standard deviation is about 14.92.

Comparing FIGS. 4A and 4B, it can be seen that the standard deviation of the pixel brightness increases in the surface reflected light image of the skin after making the matte makeup compared to the surface reflected light image of the skin before makeup. 4B and 4C, it can be seen that the standard deviation of the pixel brightness increases in the reflected light image of the surface of the skin after performing the milky makeup, when comparing the skin and the surface reflected light image after the matte makeup. Thus, in embodiments, the surface reflection light image can be used to create a glossiness index that indicates the skin's shine.

FIG. 5 is a graph showing a result of visual evaluation of skin gloss before explaining the process of generating skin gloss index according to the embodiments. The visual evaluation was carried out using a total of 60 images including an image of skin without makeup, and an image of skin with various make-up such as matte makeup, glossy make-up, and natural make-up. For a predetermined local region (e.g., a ball portion) in the image, a total of 17 subjects were visually evaluated to distinguish them as "no gloss," "gloss," or "shine." "Shine" refers to a greater degree of shine than "shine". In Fig. 5, the x-axis represents the number assigned to the image to be evaluated, and the y-axis represents the number of the experimenter evaluated as "no gloss," "glossiness" and "shine" for each image.

The glossiness index obtained by visual evaluation can be calculated by adding the scores of 0, 1, and 2 to each of the evaluation results for one image for "no glossiness", "glossiness", and "glossiness". That is, if one experimenter evaluates an arbitrary image as "no glossiness", another experimenter evaluates it as "glossy", and another experimenter evaluates it as a "glossiness" The gloss index due to the evaluation becomes 3 [= (1 x 0) + (1 x 1) + (1 x 2). Since the criteria for "no gloss", "glossiness", and "shine" are different for each experimenter, images with similar scores are obtained by adding scores of a plurality of experimenters even for different images. The rectangles 501, 502, 503, 504, and 505 shown by dotted lines in FIG. 5 represent images having similar glossiness by visual evaluation.

The glossiness index by the visual evaluation described above with reference to Fig. 5 can be used to determine one or more parameters related to generation of glossiness index according to the embodiments. This will be described in detail later.

Referring again to FIG. 1, at least one light object may be extracted from the surface reflected light image generated using the parallel polarized image and the crossed polarized image in step S3 described above (S4). In this specification, the optical object refers to a pixel having a brightness of a predetermined threshold value or more among a plurality of pixels included in a surface reflected light image. The threshold value of the pixel brightness for determining the optical object may be determined in various manners according to the embodiment.

In one embodiment, the threshold of pixel brightness for determining the optical object may be determined utilizing visual evaluation results. For example, if no image object is detected in the images (58th, 59th, and 60th images on the x axis) that do not have the experimenter evaluated by the blurred image in the visual evaluation result shown in Fig. 5 So that the threshold of pixel brightness can be determined. For example, if the maximum brightness of a pixel in an image without a shaded image is x, then (x + 1) can be determined as the threshold value of pixel brightness. The brightness of the pixel will increase in other images with relatively high degree of shine compared to the image without the shake-out evaluation, so that in other images, by setting the threshold value of the pixel brightness to (x + 1) It is possible to detect one or more optical objects having brightness equal to or greater than a threshold value.

However, the method of determining the threshold value of the pixel brightness for determining the optical object is not limited to that described above, and the threshold value may be determined by other different methods with reference to the region and characteristics of the target skin to be generated You can decide.

FIG. 6A is a surface reflected light image generated according to an embodiment, and FIG. 6B is an image representing a light object extracted from the surface reflected light image of FIG. 6A. FIG. 6B shows an enlarged view of a specific area 600 shown in FIG. For example, the region 600 may be a portion of the skin that is relatively less affected by shading due to the skin structure, such as nose or lips, but is not limited thereto. The light objects detected in the area 600 are shown in red in FIG. 6B.

Referring to FIG. 1, in order to calculate a luster index according to an exemplary embodiment, an optical object of a predetermined size or more among the detected one or more optical objects may be detected (S5). For example, of the at least one optical object made up of individual pixels, an optical object made up of a predetermined number or more of pixels connected to each other through one or more sides of the pixel can be detected. In this specification, an optical object of a predetermined size or larger determined as described above is defined as a "large" optical object. For example, an optical object composed of four or more pixels connected to each other may be detected as a "large" optical object, but is not limited thereto.

Next, the glossiness index of the skin can be calculated using the optical object having a predetermined size or more. For example, the glossiness index of the skin can be calculated using the number and / or area of the optical objects of a predetermined size or more (S6).

In one embodiment, the glossiness index is defined as the ratio of "large" optical objects in one or more optical objects. For example, the glossiness index may be a value obtained by dividing the number of "large" optical objects among the one or more optical objects by the number of optical objects excluding the "large"

Fig. 7 is a graph showing the skin gloss index calculated as above in comparison with the visual evaluation results. In the graph of FIG. 7, the x-axis represents the ratio obtained by dividing the number of "large" optical objects among the one or more optical objects by the number of remaining optical objects, and the y-axis represents the luminous index obtained by visual evaluation. As shown in the figure, it can be seen that the glossiness index according to this embodiment increases as the image is judged to have a high gloss in the visual evaluation. As a result of the regression analysis using the data shown in FIG. 7, the coefficient of determination (R ² ) between the glossiness index and the visual evaluation result according to the present embodiment was 0.7321.

In another embodiment, the glossiness index is defined as the total area occupied by "large" optical objects in one or more optical objects.

Fig. 8 is a graph showing the skin gloss index calculated as above in comparison with the visual evaluation results. In the graph of Fig. 8, the x-axis represents the area of the entire "large" optical object among the at least one optical object, and the y-axis represents the gloss index by visual evaluation. As shown in the figure, it can be seen that the glossiness index according to this embodiment increases as the image is judged to have a high gloss in the visual evaluation. As a result of the regression analysis using the data shown in FIG. 8, the coefficient of determination (R ² ) between the glossiness index and the visual evaluation result according to the present embodiment was 0.6068.

In yet another embodiment, the glossiness index is defined using both the ratio and the area of the "large" optical object in one or more optical objects. The glossiness index may be a value obtained by adding predetermined weights to the values obtained by dividing the number of "large" optical objects among the one or more optical objects by the number of remaining optical objects and the area of the entire "large" optical object. For example, if the value obtained by dividing the number of "large" optical objects by the number of remaining optical objects in one or more optical objects is X and the area of the entire "large" optical object is Y, 1 < / RTI >

[Equation 1]

I = aX + bY + c

In Equation (1), a and b are predetermined weights given to X and Y, respectively, and c is a predetermined offset constant. The weights a and b and the value of the constant c in Equation (1) can be appropriately determined in accordance with the region and the characteristics of the target skin, and are not limited to the specific values, so that the last calculated glare index is compatible with the actual condition of the skin.

In one embodiment, the weights a and b and the value of the constant c in Equation (1) may be determined through a regression analysis such that the finally calculated glossiness index is close to the glossiness index obtained by visual evaluation. For example, as a result of regression analysis of the glossiness index determined according to this embodiment and the glossiness index according to the visual evaluation shown in FIG. 5, the value of a is 0.478, the value of b is 0.000268, and the value of c is 1.49 . ≪ / RTI > A specific regression analysis method for deriving the above values is well known to those skilled in the art, so a detailed description thereof will be omitted. This regression analysis method may also be performed using known or later developed computer software.

FIG. 9 is a graph showing skin gloss index calculated by using a weight and a constant by regression analysis as compared with a visual evaluation result as described above. In the graph of FIG. 9, the x-axis represents the gloss index calculated by regression analysis with a value of 0.478, a value of b of 0.000268, and a value of c of 1.49 by the regression analysis, and the y- . As a result of the regression analysis using the data shown in FIG. 9, the coefficient of determination (R ² ) between the glossiness index and the visual evaluation result according to the present example was 0.7901.

In one embodiment, the skin luminous index calculated as described above may be divided into one or more regions according to numerical values to provide cosmetic information. In particular, the area 900 shown by the dashed line in FIG. 9 displays data in which the response of "shine" is predominant in comparison with the "shiny" in the result of the visual evaluation. Therefore, the glossiness indexes 18 to 26 corresponding to the area 900 can be determined as an index that is likely to shine, and beauty information can be provided based thereon. For example, it is possible to numerically provide the degree of the possibility of shattering when a specific cosmetic material is applied by dividing the region having a glossiness index of 18 to 26 in the first stage and the region having a glossiness index of 26 or more in the second stage have.

Table 1 shows the correlation (R) and the P-value obtained by regression analysis of the glossiness index and the visual evaluation result generated according to the embodiments, and the correlation (R) and the P- value by regression analysis of other evaluation indexes and visual evaluation results. P-value. ≪ / RTI >

Correlation with Glossiness Index by Visual Evaluation (R) P-value In a generic image
Average pixel brightness 0.387 0.002 In a generic image
Standard deviation of pixel brightness 0.834 0 In the surface reflected light image
Average pixel brightness 0.191 0.145 In the surface reflected light image
Standard deviation of pixel brightness 0.854 0 According to one embodiment,
The ratio of "large" optical objects (Figure 7) 0.856 0 According to one embodiment,
The area of the "large" light object (Figure 8) 0.753 0 According to one embodiment,
The regression equation of the ratio and area of the "large" optical object (Equation 1 / Figure 9) 0.889 0

As shown in Table 1, the brightness of the skin as a whole has a relatively small influence on the glossiness, and the brightness variation shows a relatively large correlation with the glossiness. According to the embodiments, since the glossiness index is calculated and analyzed according to the ratio and / or area of the "large" optical object among the at least one optical object, the glossiness index is highly correlated with the result of the visual evaluation, And the glossiness index is in good agreement with the actual glossiness of the skin.

The above-described method of generating skin luster indexes has been described with reference to the flowcharts shown in the drawings. While the above method has been shown and described as a series of blocks for purposes of simplicity, it is to be understood that the invention is not limited to the order of the blocks, and that some blocks may be present in different orders and in different orders from that shown and described herein And various other branches, flow paths, and sequences of blocks that achieve the same or similar results may be implemented. Also, not all illustrated blocks may be required for implementation of the methods described herein.

While the invention has been shown and described with reference to certain preferred embodiments thereof, it will be understood by those skilled in the art that various changes and modifications may be made therein without departing from the scope of the invention as defined by the appended claims. will be. However, it should be understood that such modifications are within the technical scope of the present invention. Accordingly, the true scope of the present invention should be determined by the technical idea of the appended claims.

Claims (6)

Photographing a cross-polarized image of the skin by the camera;
Photographing a parallel polarized image of the skin by the camera;
Obtaining, by the computing device, a surface reflected light image corresponding to a difference between the parallel and cross polarization images;
Detecting, by the computing device, one or more optical objects from the surface reflected light image; And
Calculating a glossiness index by using the optical object of a predetermined size or larger among the one or more optical objects by the computing device,
Wherein detecting the at least one optical object comprises detecting a pixel having a brightness greater than or equal to a predetermined threshold value in the surface reflected light image.
delete The method according to claim 1,
The step of calculating the luster index may further include calculating a value obtained by dividing the number of the optical objects having the predetermined size or more by the number of the optical objects excluding the optical object having the predetermined size or more among the one or more optical objects Wherein the skin luster index is generated by the skin luster index.
The method according to claim 1 or 3,
Wherein the step of calculating the glare index includes calculating an area of the optical object of the predetermined size or larger.
5. The method of claim 4,
Wherein the step of calculating the glossiness index further comprises the step of summing and adding predetermined weights to each of the value and the area.
6. The method of claim 5,
Wherein the weighting and summing comprises determining the weight by regression analysis with a visual evaluation result.

Images