JP7796483B2 - Methods for evaluating the degree of eyelash variation, methods for evaluating the degree of eyelash curl, impression evaluation methods, counseling methods, and market research methods - Google Patents

Description

The present invention relates to a method for evaluating the degree of eyelash variation, a method for evaluating the degree of eyelash curl, a method for evaluating impressions, a counseling method, and a market research method based on a side view of the eye.

The eyes are a feature that has a significant impact on the impression one makes on others. In the current COVID-19 pandemic, people are required to wear masks on a regular basis, making the impression of the eyes even more important. Furthermore, with people now being encouraged to avoid face-to-face communication, the impression of the eyes in profile has become even more important.

The cosmetics industry is seeking to develop new eyelash preparations and propose makeup methods suited to this new normal era, and eye impression evaluation methods will be an important tool for such development and proposals.

Among the studies on methods for evaluating eye impressions to date, for example, Patent Document 1 describes a method in which a line segment BH connecting the base B and tip H of an eyelash sample taken from a subject is used (see Patent Document 1). Furthermore, for example, Patent Document 2 describes a method in which mascara is applied to a customer's eyelashes, the customer's posture is such that their eyes are facing horizontally, and the curl angle is measured, which is the angle between the horizontal line and the tangent to the base of the curled eyelashes (see Patent Document 2).

Japanese Patent Application Laid-Open No. 2019-137647 WO 2002/051285

However, the above-mentioned conventional methods for evaluating eye impressions leave room for improvement in terms of accuracy and ease of use.

Therefore, an object of the present invention is to provide a method for accurately and simply evaluating the degree of curl and the degree of variation of eyelashes.
Another object of the present invention is to provide a method for evaluating an impression of a subject based on the degree of eyelash curl and the degree of eyelash variation.A further object of the present invention is to provide a method for providing counseling to a subject based on the degree of eyelash curl and the degree of eyelash variation.A further object of the present invention is to investigate the correlation between the degree of eyelash curl and the degree of eyelash variation and evaluations of each impression evaluation item included in a group of impression evaluation words.

The gist of the present invention is as follows.
[1]
A method for evaluating the degree of eyelash variation, which calculates an unbiased variance V(θ) of the angle θ between a line segment BH connecting the base B and tip H of a plurality of eyelashes and a horizontal line HL in a side image of the subject's eye, and evaluates the degree of variation of the subject's eyelashes based on the unbiased variance V(θ).
[2]
The method for evaluating the degree of eyelash variation according to [1], wherein the plurality of eyelashes are eyelashes whose bases B are located at the top of the eye.
[3]
The method for evaluating the degree of eyelash variation according to [1] or [2], wherein the angle θ of all of the plurality of eyelashes is in the range of −40° to +40° (horizon is taken as 0°), and when all of the eyelashes are ranked from the smallest angle θ to the largest angle θ and the difference in angle θ between the ranked positions is calculated as an absolute value, the average of the differences is 10° or less.
[4]
The method for evaluating the degree of eyelash variation according to any one of [1] to [3], wherein the plurality of eyelashes are eyelashes located in the center of the upper eyelid in a frontal eye image of the subject.
[5]
A method for evaluating the degree of eyelash curl, comprising: calculating a curl value S/Le by dividing an area S of a region enclosed by a line segment BH connecting the base B and tip H of the eyelash and a line tracing the line from the base B to the tip H of the eyelash in a side view image of the eye of a subject, by a length Le of the line tracing the line from the base B to the tip H of the eyelash; and evaluating the degree of curl of the eyelash based on the curl value S/Le.
[6]
The method for measuring the degree of curl of eyelashes according to [5], wherein the degree of curl of eyelashes is evaluated based on an average value of the curl values S/Le calculated for a plurality of eyelashes.
[7]
The method for evaluating the degree of curl of eyelashes according to [5] or [6], wherein the base B of the plurality of eyelashes is the eyelashes located at the top of the eye in the side view eye image.
[8]
The method for evaluating the degree of curl of eyelashes according to any one of [5] to [7], wherein the plurality of eyelashes are those having the first to third longest lengths of the line segment BH among the eyelashes located at the top of the eye, and the angle θ of all the eyelashes is in the range of −40° to +40° (horizon is taken as 0°), and when all the eyelashes are ranked from the smallest angle θ to the largest angle θ and the difference in angle θ between the ranked eyelashes is calculated as an absolute value, the average of the differences is 10° or less.
[9]
The method for evaluating the degree of eyelash variation described in [8], wherein the plurality of eyelashes are eyelashes located in the center of the upper eyelid in the frontal eye image of the subject.
[10]
In the side view image of the subject's eye, the unbiased variance V(θ) of the angle θ between the horizontal line HL and the line segment BH connecting the base B and tip H of the eyelashes calculated for multiple eyelashes.
and/or
In the side view of the eye of the subject, the area S of the region surrounded by the line segment BH connecting the base B and the tip H of the eyelash and the line tracing the line from the base B to the tip H of the eyelash is divided by the length Le of the line tracing the line from the base B to the tip H of the eyelash, to obtain a curl value S/Le;
and performing an impression evaluation on the subject based on the above.
[11]
preparing a plurality of side view images of the eye having different unbiased variances V(θ) and/or different curl values S/Le;
The panelists evaluate the plurality of side eye images with respect to each impression evaluation item included in the impression evaluation word group,
creating reference data indicating a correlation between the unbiased variance V(θ) and/or the curl value S/Le and an evaluation regarding each of the impression evaluation items included in the group of impression evaluation words;
performing an impression evaluation on the subject using the reference data;
The impression evaluation method according to [10].
[12]
The impression evaluation method described in [11], wherein the impression evaluation is performed by performing an absolute evaluation for each impression evaluation item included in the group of impression evaluation words for one or more side view eye images based on the reference data.
[13]
The impression evaluation method according to [12], wherein the subject's impression is classified into types based on the absolute evaluation.
[14]
preparing a plurality of side view images of the eye having different unbiased variances V(θ) and/or different curl values S/Le;
For any pair of the side eye images among the plurality of side eye images,
having the panelists compare and evaluate one of the side view images of the eyes with the other side view image of the eyes with respect to each of the impression evaluation items included in the group of impression evaluation words;
creating reference data indicating a correlation between a difference between the unbiased variance V(θ) and/or the curl value S/Le in the one side view eye image and the unbiased variance V(θ) and/or the curl value S/Le in the other side view eye image, and a comparative evaluation of each of the impression evaluation items included in the group of impression evaluation words;
performing an impression evaluation on the subject using the reference data;
The impression evaluation method according to [10].
[15]
The impression evaluation method according to [14], wherein the impression evaluation is performed by comparatively evaluating a plurality of side view eye images with respect to each impression evaluation item included in the group of impression evaluation words based on the reference data.
[16]
The impression evaluation method according to any one of [10] to [15], wherein the reference data is updated periodically or irregularly.
[17]
The impression evaluation method according to any one of [10] to [16], wherein the panelists are at least one selected from the group consisting of experts, researchers, beauty consultants, and consumers.
[18]
In the side view image of the subject's eyes, the unbiased variance V(θ) of the angle θ between the horizontal line HL and the line segment BH connecting the base B and tip H of the eyelashes calculated for multiple eyelashes,
and/or
In the side view of the subject's eyes, the area S of the region surrounded by the line segment BH connecting the base B and the tip H of the eyelashes and the line tracing the line from the base B to the tip H of the eyelashes is divided by the length Le of the line tracing the line from the base to the tip of the eyelashes to obtain the curl value S/Le.
A counseling method in which counseling is provided to a subject based on the above.
[19]
The counseling method described in [18] includes modifying a side view image of the subject's eyes to change the degree of eyelash variation and/or curl degree by image modification, thereby obtaining a side view simulation image of the eyes having the unbiased variance V(θ) and/or the curl value S/Le that are different from those before the modification.
[20]
The counseling method according to [19], further comprising making an impression evaluation of the side eye simulation image.
[21]
preparing a plurality of side view eye simulation images having different unbiased variances V(θ) and/or different curl values S/Le;
The panelists evaluate the plurality of side eye simulation images with respect to each impression evaluation item included in the impression evaluation word group,
creating reference data indicating a correlation between the unbiased variance V(θ) and/or the curl value S/Le and an evaluation regarding each of the impression evaluation items included in the group of impression evaluation words;
Using the reference data, an impression evaluation is performed on the subject after the simulation.
The counseling method described in [20].
[22]
The counseling method described in [21], wherein the impression evaluation involves performing an absolute evaluation of one or more of the side eye simulation images with respect to each impression evaluation item included in the group of impression evaluation words based on the reference data.
[23]
The counseling method according to [22], wherein the subject's impression after the simulation is classified into types based on the absolute evaluation.
[24]
preparing a plurality of side view eye simulation images having different unbiased variances V(θ) and/or different curl values S/Le;
For any pair of side view eye simulation images among the plurality of side view eye simulation images,
having panelists compare and evaluate one of the side view eye simulation images with the other side view eye simulation image with respect to each of the impression evaluation items included in the group of impression evaluation words;
creating reference data indicating a correlation between a difference between the unbiased variance V(θ) and/or the curl value S/Le in one of the side-view eye simulation images and the unbiased variance V(θ) and/or the curl value S/Le in the other of the side-view eye simulation images, and a comparative evaluation of each of the impression evaluation items included in the group of impression evaluation words;
Using the reference data, an impression evaluation is performed on the subject after the simulation.
The counseling method described in [20].
[25]
The counseling method according to [24], wherein the impression evaluation involves comparative evaluation of a plurality of side view eye simulation images with respect to each impression evaluation item included in the group of impression evaluation words based on the reference data.
[26]
The counseling method according to any one of [20] to [25], wherein the content of a suggestion regarding eyelash cosmetics and/or eyelash makeup methods is determined according to the result of the impression evaluation.
[27]
The counseling method according to [18], further comprising applying eyelash cosmetics and/or performing an eyelash makeup method on the side image of the subject's eyes to obtain a side image of made-up eyes.
[28]
The counseling method according to [27], further comprising carrying out the impression evaluation on the side eye makeup image.
[29]
preparing a plurality of side view eye makeup images having different unbiased variances V(θ) and/or different curl values S/Le;
The panelists evaluate the plurality of side eye makeup images with respect to each impression evaluation item included in the group of impression evaluation words,
creating reference data indicating a correlation between the unbiased variance V(θ) and/or the curl value S/Le and an evaluation regarding each of the impression evaluation items included in the group of impression evaluation words;
Using the reference data, an impression evaluation is performed on the subject after makeup is applied.
The counseling method described in [28].
[30]
The counseling method described in [29], wherein the impression evaluation involves performing an absolute evaluation of one or more of the side eye makeup images with respect to each impression evaluation item included in the group of impression evaluation words based on the reference data.
[31]
The counseling method according to [30], wherein the impression of the subject after makeup is classified into types based on the absolute evaluation.
[32]
preparing a plurality of side view eye makeup images having different unbiased variances V(θ) and/or different curl values S/Le;
Regarding any pair of side eye makeup images among the plurality of side eye makeup images,
having panelists compare and evaluate one of the side view eye makeup images with the other side view eye makeup image with respect to each of the impression evaluation items included in the group of impression evaluation words;
creating reference data indicating a correlation between a difference between the unbiased variance V(θ) and/or the curl value S/Le in one of the side eye makeup images and the unbiased variance V(θ) and/or the curl value S/Le in the other side eye makeup image, and a comparative evaluation of each of the impression evaluation items included in the group of impression evaluation words;
Using the reference data, an impression evaluation is performed on the subject after the simulation.
The counseling method described in [27].
[33]
A counseling method for providing counseling to a subject using the method for evaluating the degree of eyelash variation according to any one of [1] to [4], the method for measuring the degree of eyelash curl according to any one of [5] to [9], or the method for evaluating the impression according to any one of [10] to [17].
[34]
In the side view image of the subject's eyes, the unbiased variance V(θ) of the angle θ between the horizontal line HL and the line segment BH connecting the base B and tip H of the eyelashes calculated for multiple eyelashes,
and/or
In the side view of the subject's eyes, the area S of the region surrounded by the line segment BH connecting the base B and the tip H of the eyelashes and the line tracing the line from the base B to the tip H of the eyelashes is divided by the length Le of the line tracing the line from the base to the tip of the eyelashes to obtain the curl value S/Le.
Prepare a plurality of side view images of the eye with different
A correlation between the unbiased variance V(θ) and/or the curl value S/Le and the evaluation of each of the impression evaluation items included in the group of impression evaluation words is investigated by having panelists evaluate the plurality of side view eye images with respect to each of the impression evaluation items included in the group of impression evaluation words.
Market research methods.
[35]
In the side view image of the subject's eyes, the unbiased variance V(θ) of the angle θ between the line segment BH connecting the base B and tip H of the eyelash and the horizontal line HL, calculated for multiple eyelashes,
and/or
In the side view of the subject's eyes, the area S of the region surrounded by the line segment BH connecting the base B and the tip H of the eyelashes and the line tracing the line from the base B to the tip H of the eyelashes is divided by the length Le of the line tracing the line from the base to the tip of the eyelashes to obtain the curl value S/Le.
Prepare a plurality of side view images of the eye with different
For any pair of the side eye images among the plurality of side eye images,
having the panelists compare and evaluate one of the side view images of the eyes with the other side view image of the eyes with respect to each of the impression evaluation items included in the group of impression evaluation words;
investigating a correlation between a difference between the unbiased variance V(θ) and/or the curl value S/Le in the one side view eye image and the unbiased variance V(θ) and/or the curl value S/Le in the other side view eye image and a comparative evaluation of each of the impression evaluation items included in the impression evaluation word group;
Market research methods.
[36]
In the side view image of the subject's eyes, the unbiased variance V(θ) of the angle θ between the horizontal line HL and the line segment BH connecting the base B and tip H of the eyelashes calculated for multiple eyelashes,
and/or
In the side view of the subject's eyes, the area S of the region surrounded by the line segment BH connecting the base B and the tip H of the eyelashes and the line tracing the line from the base B to the tip H of the eyelashes is divided by the length Le of the line tracing the line from the base to the tip of the eyelashes to obtain the curl value S/Le.
A plurality of side eye images with different
an average side view eye image obtained by averaging the plurality of side view eye images having different unbiased variances V(θ) and/or different curl values S/Le;
Prepare
having the panelists compare each of the side eye images with the average side eye image, thereby having the panelists evaluate the plurality of side eye images with respect to each impression evaluation item included in a group of impression evaluation words;
Market research methods.
[37]
The market research method according to any one of [32] to [34], wherein the panelists are at least one selected from the group consisting of experts, researchers, beauty consultants, and consumers.

According to the present invention, it is possible to provide a method for accurately and simply evaluating the degree of curl and the degree of variation of eyelashes.
The present invention also provides a method for evaluating an impression of a subject based on the degree of eyelash curl and the degree of eyelash variation. Furthermore, the present invention also provides a method for providing counseling to a subject based on the degree of eyelash curl and the degree of eyelash variation. Furthermore, the present invention also makes it possible to investigate the correlation between the degree of eyelash curl and the degree of eyelash variation and the evaluation of each impression evaluation item included in a group of impression evaluation words.

Fig. 1 is a diagram showing a schematic side view of the eye of a subject (subject). In Fig. 1, only one eyelash of the subject to be evaluated is shown. Fig. 2(A) is a diagram schematically showing the apex of the eye as determined in a side view image of the subject's (target's) eyes. In Fig. 2(A), the horizontal outermost end of the edge of the upper eyelid is indicated by an "x", and the apex of the eye is indicated by a dashed circle. Fig. 2(B) is a diagram schematically showing the center of the upper eyelid as determined in a front view image of the subject's (target's) eyes. In Fig. 2(B), the midpoint of the line tracing the upper eyelid from its start (inner corner) to its end (outer corner) is indicated by an "x", and the center of the upper eyelid is indicated by a dashed circle. 3A and 3B are diagrams illustrating a state in which a side view image of the eyes of a subject (target person) is acquired, respectively, and a front view image of the eyes of the subject (target person).

Hereinafter, with reference to the drawings, embodiments of the method for evaluating the degree of eyelash variation of the present invention, the method for evaluating the degree of eyelash curl of the present invention, the impression evaluation method of the present invention, the counseling method of the present invention, and the market research method of the present invention will be described in detail.
Hereinafter, in this specification, an embodiment of the present invention will also be referred to as "the present embodiment."

(Method for evaluating the degree of eyelash variation)
The method for evaluating the degree of eyelash variation in this embodiment calculates the unbiased variance V(θ) of the angle θ between the line segment BH connecting the base B and tip H of each eyelash and the horizontal line HL for multiple eyelashes in a side view image of the subject's eye, and evaluates the degree of eyelash variation of the subject based on the unbiased variance V(θ).

The side view image of the subject's eyes is not particularly limited and may be obtained by taking a photograph using a general imaging device from the side of the subject, more specifically from a direction along the line connecting the tops of the left and right ears, and specifically may be obtained using the method described in the examples.

Figure 1 is a schematic diagram showing a side view of the subject's (subject's) eyes. In Figure 1, only one eyelash of the subject (subject's) that is being evaluated is shown.

The unbiased variance V(θ) is calculated using the following formula (1):

...(1)

In equation (1), θi represents the angle θ for the i-th eyelash, and AVE(θ) represents the average angle θ for n eyelashes.

In this embodiment, the plurality of eyelashes used to calculate the unbiased variance V(θ) are not particularly limited and may be selected from all eyelashes included in the side view eye image. The plurality of eyelashes may be eyelashes whose bases B are located at the top of the eye or eyelashes whose bases B are located between the top of the eye and the outer corner of the eye, and it is particularly preferable to select eyelashes whose bases B are located at the top of the eye.
The top of the eye refers to the area within 2 mm from the outermost horizontal edge of the upper eyelid (see the "x" mark in Figure 2(A)) in the side view of the eye (see the dashed circle in Figure 2(A)).

The following describes a preferred embodiment of a method for evaluating the degree of eyelash variation in which eyelashes whose base B is located at the top of the eye are evaluated, but the same applies to other embodiments in which eyelashes located in other locations are evaluated.

In a preferred embodiment, all eyelashes at the top of the eye may be evaluated, or a representative selection of eyelashes may be evaluated. It is preferable to select the representative selection of eyelashes such that the unbiased variance V(θ) for the selected selection of eyelashes falls within an error range of -20% to +20% based on the unbiased variance V(θ) for all eyelashes.

The angle θ for all the eyelashes at the top of the eye generally tends to be distributed almost uniformly and without significant deviation within the range of −50° to +50° (the horizontal line being 0°).
In view of this tendency, it is preferable, from the viewpoint of improving the accuracy of the evaluation, that the angle θ of all of the multiple eyelashes selected for evaluation be in the range of −40° to +40° (with the horizontal line being 0°), preferably in the range of −30° to +30° (with the horizontal line being 0°), and that when all of the eyelashes are ranked from smallest to largest angle θ and the difference in angle θ between the ranked lines is calculated as an absolute value, the average of the difference be 10° or less, preferably 7.5° or less, and even more preferably 7.5° or less.

From the perspective of increasing the accuracy of the evaluation, it is preferable to select the multiple eyelashes to be evaluated from among the eyelashes located at the top of the eye that have the first, second, and third longest line segment BH lengths.

In this embodiment, the plurality of eyelashes are preferably eyelashes located in the center of the upper eyelid in the frontal eye image of the subject.
The central part of the upper eyelid refers to the area within 2 mm from the midpoint of the line tracing the upper eyelid from its beginning (inner corner of the eye) to its end (outer corner of the eye) in the frontal eye image (see the "x" mark in Figure 2(B)).

The frontal eye image of the subject is not particularly limited and may be obtained by photographing the subject from the front, more specifically from a direction perpendicular to the line connecting the tops of the left and right ears and along the horizontal line, using a general imaging device; specifically, it may be obtained using the method described in the examples.

The frontal eye image may be used as an auxiliary means for selecting eyelashes suitable for evaluation in the side eye image.
For example, when selecting an eyelash whose base B is located at the apex of the eye in the side view eye image, it may be difficult to select the eyelash because the bases of other eyelashes are present near the base B. In this case, it is possible to select multiple eyelashes to be evaluated by selecting an eyelash as an evaluation candidate from among the eyelashes located at the center of the upper eyelid in the front view eye image, and then checking in the side view eye image whether the selected eyelash as an evaluation candidate has its base B located at the apex of the eye.

The method for evaluating the degree of eyelash variation of this embodiment can be applied not only to cases where a side view of the eyes is used, but also to cases where an image obtained by photographing the face from an oblique front or an oblique rear view is used.

(Method for evaluating the degree of eyelash curl)
The method for evaluating the degree of eyelash curl in this embodiment calculates a curl value S/Le by dividing the area S of the region surrounded by a line segment BH connecting the base B and tip H of the eyelash and a line tracing from the base B to the tip H of the eyelash in a side view image of the subject's eye by the length Le of the line tracing from the base B to the tip H of the eyelash, and evaluates the degree of eyelash curl based on the curl value S/Le.
The side view image of the subject's eyes is not particularly limited and may be obtained by photographing the subject from the side, more specifically from a direction along the line connecting the top ends of the left and right ears, using a general imaging device, and specifically may be obtained using the method described in the examples.

Figure 1 is a schematic diagram showing a side view of the subject's (subject's) eyes. In Figure 1, only one eyelash of the subject (subject's) that is being evaluated is shown.

The curl value S/Le is calculated by using image analysis to calculate the area S of the region surrounded by the line segment BH connecting the base B and tip H of the eyelash and the line tracing the line from the base B to the tip H of the eyelash, and then using image analysis to calculate the length Le of the line tracing the line from the base B to the tip H of the eyelash, and dividing the calculated area S by the calculated Le.

In this embodiment, it is preferable to evaluate the degree of curl of the eyelashes based on the average value of the curl values S/Le calculated for multiple eyelashes. By using the average value, it is possible to evaluate the entire selected multiple eyelashes.

In this embodiment, it is preferable that the bases B of the eyelashes are located at the top of the eye in the side view of the eye.
The top of the eye refers to the area within 2 mm from the outermost horizontal edge of the upper eyelid (see the "x" mark in Figure 2(A)) in the side view of the eye (see the dashed circle in Figure 2(A)).

The following describes a preferred embodiment of a method for evaluating the degree of curl of eyelashes, in which the base B of the eyelashes is located at the top of the eye, but the same applies to other embodiments in which eyelashes located in other positions are evaluated.

In a preferred embodiment, all eyelashes at the top of the eye may be evaluated, or a representative selection of eyelashes may be evaluated. It is preferable to select the representative selection of eyelashes such that the unbiased variance V(θ) for the selected selection of eyelashes falls within an error range of -20% to +20% based on the unbiased variance V(θ) for all eyelashes.

The angle θ for all the eyelashes at the top of the eye generally tends to be distributed almost uniformly and without significant deviation within the range of −50° to +50° (the horizontal line being 0°).
In view of this tendency, it is preferable, from the viewpoint of improving the accuracy of the evaluation, that the angle θ of all of the multiple eyelashes selected for evaluation be in the range of −40° to +40° (with the horizontal line being 0°), preferably in the range of −30° to +30° (with the horizontal line being 0°), and that when all of the eyelashes are ranked from smallest to largest angle θ and the difference in angle θ between the ranked lines is calculated as an absolute value, the average of the difference be 10° or less, preferably 7.5° or less, and even more preferably 7.5° or less.

From the perspective of increasing the accuracy of the evaluation, it is preferable to select the multiple eyelashes to be evaluated from among the eyelashes located at the top of the eye that have the first, second, and third longest line segment BH lengths.

In this embodiment, the eyelashes are those located in the center of the upper eyelid in the frontal eye image of the subject.
The frontal eye image of the subject is not particularly limited and may be obtained by photographing the subject from the front, more specifically from a direction perpendicular to the line connecting the tops of the left and right ears and along the horizontal line, using a general imaging device, and specifically may be obtained by the method described in the examples.

The frontal eye image may be used as an auxiliary means for selecting eyelashes suitable for evaluation in the side eye image.
For example, when selecting an eyelash whose base B is located at the apex of the eye in the side view eye image, it may be difficult to select the eyelash because the bases of other eyelashes are present near the base B. In this case, it is possible to select multiple eyelashes to be evaluated by selecting an eyelash as an evaluation candidate from among the eyelashes located at the center of the upper eyelid in the front view eye image, and then checking in the side view eye image whether the selected eyelash as an evaluation candidate has its base B located at the apex of the eye.

The central part of the upper eyelid refers to the area within 2 mm of the midpoint of the line tracing the upper eyelid from its beginning (inner corner) to its end (outer corner) in the frontal eye image (see the "x" mark in Figure 2(B)). (See the dashed circle in Figure 2(B)).

The method for evaluating the degree of eyelash curl of this embodiment can be applied not only to cases where a side view of the eyes is used, but also to cases where an image obtained by photographing the face from an oblique front or an oblique rear view is used.

(Impression evaluation method)
The impression evaluation method of this embodiment includes:
In the side view image of the subject's eye, the unbiased variance V(θ) of the angle θ between the horizontal line HL and the line segment BH connecting the base B and tip H of the eyelashes calculated for multiple eyelashes.
and/or
In the side view of the eye of the subject, the area S of the region surrounded by the line segment BH connecting the base B and the tip H of the eyelash and the line tracing the line from the base B to the tip H of the eyelash is divided by the length Le of the line tracing the line from the base B to the tip H of the eyelash, to obtain the curl value S/Le.
This is a method for evaluating the impressions of subjects based on the above.

Specifically, in this embodiment, a plurality of side view images of the eye having different unbiased variances V(θ) and/or curl values S/Le are prepared,
The panelists evaluate the plurality of side eye images with respect to each impression evaluation item included in the impression evaluation word group,
creating reference data (hereinafter also referred to as "first reference data") that indicates a correlation between the unbiased variance V(θ) and/or the curl value S/Le and an evaluation regarding each of the impression evaluation items included in the group of impression evaluation words;
performing an impression evaluation on the subject using the reference data (first reference data);
It is preferable.
Here, when preparing multiple side view eye images having different unbiased variances V(θ) and/or curl values S/Le, side view eye images of multiple subjects having different unbiased variances V(θ) and/or curl values S/Le may be prepared, or multiple side view eye images having different unbiased variances V(θ) and/or curl values S/Le may be prepared by performing image correction on the side view eye image of one subject having a specific unbiased variance V(θ) and/or curl value S/Le.
Image correction may be performed using commercially available or free image analysis software (e.g., ImageJ, etc.).

The impression evaluation items included in the group of impression evaluation words are not particularly limited and may be selected appropriately depending on the purpose and use, but examples include intelligent, beautiful, motivated, proactive, decisive, gentle, honest, cheerful, friendly, nice, strong-willed, dignified, aesthetically pleasing, sophisticated, confident, approachable, clean, sociable, curious, and calm.
These impression evaluation items may be used alone or in combination of two or more to form the group of impression evaluation words.
The evaluation of each impression evaluation item may be performed qualitatively or quantitatively. When the evaluation is performed quantitatively, the evaluation may be performed by assigning a score (for example, from −3 to +3).

The reference data (first reference data) here may indicate the correlation between the unbiased variance V(θ) and the evaluation for each impression evaluation word included in the impression evaluation word group, or may indicate the correlation between the curl value S/Le and the evaluation for each impression evaluation word included in the impression evaluation word group, or may indicate the correlation between the unbiased variance V(θ) and the curl value S/Le and the evaluation for each impression evaluation word included in the impression evaluation word group.
The reference data (first reference data) is preferably based on two or more side view images of the eyes, more preferably four or more side view images of the eyes, and even more preferably six or more side view images of the eyes.
Here, the correlation is not particularly limited, and may be a positive correlation, a negative correlation, or a relationship that provides reasonable predictability. The strength of the correlation is also not particularly limited.

In the impression evaluation in this embodiment, it is preferable to perform an absolute evaluation of one or more side view eye images with respect to each impression evaluation item included in the group of impression evaluation words based on reference data (first reference data).
Absolute evaluation may be performed, for example, by calculating the unbiased variance V(θ) for one or more side view eye images to be evaluated, and then determining the score for each impression evaluation item to be assigned to the calculated unbiased variance V(θ) in the reference data (first reference data).
For example, if V(θ) is calculated as 100 for the first side view eye image, and V(θ) is 20, 60, 140, or 200 in the reference data (first reference data), and the scores for the "intelligent" item are assigned as -2, -1, +1, or +2, respectively, the score for the "intelligent" item for the first side view eye image can be calculated as 0.

Furthermore, in this embodiment, it is preferable to classify the subject's impressions into types based on absolute evaluation.
Regarding the type used for classification, for example, an impression that is evaluated with a specific range of scores for the first impression evaluation item, a specific range of scores for the second impression evaluation item, a specific range of scores for the third impression evaluation item, and a specific range of scores for the fourth impression evaluation item can be set as a specific type.
For example, an impression that is evaluated as +1 to +3 for the "intelligent" item, +1 to +3 for the "decisive" item, +1 to +3 for the "refined" item, and +1 to +3 for the "clean" item can be considered to be an elegant type.

In the impression evaluation method of this embodiment, the above-mentioned types, which are composed of the types of impression evaluation items and the ranges of ratings, may be updated regularly or irregularly. By updating regularly or irregularly, it becomes possible to adjust the types to suit the times, social climate, trends, etc.

Specifically, in this embodiment, a plurality of side view images of the eye having different unbiased variances V(θ) and/or different curl values S/Le are prepared,
For any pair of the side eye images among the plurality of side eye images,
having the panelists compare and evaluate one of the side view images of the eyes with the other side view image of the eyes with respect to each of the impression evaluation items included in the group of impression evaluation words;
creating reference data (hereinafter also referred to as "second reference data") that indicates a correlation between a difference between the unbiased variance V(θ) and/or the curl value S/Le in the one side view eye image and the unbiased variance V(θ) and/or the curl value S/Le in the other side view eye image, and a comparative evaluation of each of the impression evaluation items included in the group of impression evaluation words;
performing an impression evaluation on the subject using the reference data (second reference data);
It is preferable.

In the impression evaluation in this embodiment, it is preferable to perform a comparative evaluation of a plurality of side view eye images with respect to each impression evaluation item included in the group of impression evaluation words based on reference data (second reference data).
The comparative evaluation may be performed, for example, by calculating the unbiased variance V(θ) of any pair of side eye images among the multiple side eye images to be evaluated, then calculating the difference between the unbiased variances V(θ) of the two, and determining the score for each impression evaluation item of one side eye image in comparison with the other side eye image for the pair of side eye images in comparison with the other side eye image (the score for each impression evaluation item included in the impression evaluation word group assigned to one side eye image when one side eye image is evaluated in comparison with the other side eye image).
For example, a pair of a first side view eye image and a second side view eye image is prepared, and V(θ) is calculated as 10 for the first side view eye image and V(θ) as 110 for the second side view eye image. If the difference in V(θ) is 20, 60, 140, or 200 in the reference data (second reference data), and the scores for the "intelligent" category are assigned as -2, -1, +1, or +2, respectively, the score for the "intelligent" category in comparison between the first side view eye image and the second side view eye image can be calculated as 0.

In the impression evaluation method of this embodiment, it is preferable to update the reference data (second reference data) regularly or irregularly. By updating regularly or irregularly, it becomes possible to adjust the reference data (second reference data) to suit the times, social climate, trends, etc.

In the impression evaluation method of this embodiment, it is preferable that the panelists be at least one selected from the group consisting of experts, researchers, beauty consultants, and consumers.

The impression evaluation method of this embodiment can be applied not only to cases where a side view of the eyes is used, but also to cases where an image obtained by photographing the face from an oblique front or an oblique rear view is used.

(Counseling methods)
The counseling method of this embodiment includes:
In the side view image of the subject's eyes, the unbiased variance V(θ) of the angle θ between the horizontal line HL and the line segment BH connecting the base B and tip H of the eyelashes calculated for multiple eyelashes,
and/or
In the side view of the subject's eyes, the area S of the region surrounded by the line segment BH connecting the base B and the tip H of the eyelashes and the line tracing from the base B to the tip H of the eyelashes is divided by the length Le of the line tracing from the base to the tip of the eyelashes to obtain the curl value S/Le.
This is a method of providing counseling to subjects based on the above.

Specifically, the counseling method of this embodiment includes modifying a side view image of the subject's eyes to change the degree of eyelash variation and/or curl by image modification, thereby obtaining a side view simulation image of the eyes that has a different unbiased variance V(θ) and/or curl value S/Le from before the modification.
Here, when obtaining a side view eye simulation image having an unbiased variance V(θ) and/or curl value S/Le that is different from that before the image modification, one or more specific eyelashes of the subject may be deleted or modified, or one or more virtual eyelashes may be added to the subject.
The side view eye simulation image may be created assuming that eyelash cosmetics and/or eyelash makeup techniques are applied.

The counseling method of this embodiment includes performing an impression evaluation on a side view eye simulation image.

The impression evaluation of the side view eye simulation image here may be the same as the impression evaluation embodiment in the impression evaluation method of this embodiment (see above), except for embodiments specific to the counseling method of this embodiment.

In the impression evaluation in the counseling method of the present embodiment, a plurality of side-view eye simulation images having different unbiased variances V(θ) and/or different curl values S/Le are prepared,
The panelists evaluate the plurality of side eye simulation images with respect to each impression evaluation item included in the impression evaluation word group,
creating reference data (first reference data) that indicates a correlation between the unbiased variance V(θ) and/or the curl value S/Le and an evaluation regarding each of the impression evaluation items included in the group of impression evaluation words;
It is preferable to use the reference data to perform an impression evaluation on the subject after the simulation.

In the impression evaluation in the counseling method of this embodiment, it is preferable to perform an absolute evaluation of one or more side view eye simulation images with respect to each impression evaluation item included in the group of impression evaluation words based on reference data (first reference data).
Absolute evaluation may be performed, for example, by calculating the unbiased variance V(θ) for one or more side-view eye simulation images to be evaluated, and then determining the score for each impression evaluation item to be assigned to the calculated unbiased variance V(θ) in the reference data (first reference data).
For example, if V(θ) = 100 is calculated for the first side view eye simulation image, and V(θ) = 20, 60, 140, or 200 in the reference data (first reference data), and the scores for the "intelligent" item are assigned -2, -1, +1, or +2, respectively, then the score for the "intelligent" item for the first side view eye simulation image can be calculated as 0.

Furthermore, in this embodiment, it is preferable to classify the subject's impression after the change due to image correction into types based on absolute evaluation.
Regarding the type used for classification, for example, an impression that is evaluated with a specific range of scores for the first impression evaluation item, a specific range of scores for the second impression evaluation item, a specific range of scores for the third impression evaluation item, and a specific range of scores for the fourth impression evaluation item can be set as a specific type.
For example, an impression that is evaluated as +1 to +3 for the "intelligent" item, +1 to +3 for the "decisive" item, +1 to +3 for the "refined" item, and +1 to +3 for the "clean" item can be considered to be an elegant type.

In the impression evaluation in the counseling method of the present embodiment, a plurality of side-view eye simulation images having different unbiased variances V(θ) and/or different curl values S/Le are prepared,
For any pair of side view eye simulation images among the plurality of side view eye simulation images,
having the panelists compare and evaluate one of the side view eye simulation images with the other side view eye simulation image with respect to each of the impression evaluation items included in the group of impression evaluation words;
creating reference data (second reference data) that indicates a correlation between a difference between the unbiased variance V(θ) and/or the curl value S/Le in one of the side-view eye simulation images and the unbiased variance V(θ) and/or the curl value S/Le in the other of the side-view eye simulation images, and a comparative evaluation of each of the impression evaluation items included in the group of impression evaluation words;
It is preferable to use the reference data (second reference data) to perform an impression evaluation on the subject after the simulation.

In the impression evaluation in the counseling method of this embodiment, it is preferable to perform a comparative evaluation of multiple side view eye simulation images with respect to each impression evaluation item included in the group of impression evaluation words based on reference data (second reference data).
The comparative evaluation may be performed, for example, by calculating the unbiased variance V(θ) of any pair of side view eye simulation images among the multiple side view eye simulation images to be evaluated, then calculating the difference between the unbiased variances V(θ) of both, and determining the score for each impression evaluation item of one side view eye simulation image in comparison with the other side view eye simulation image (the score for each impression evaluation item included in the impression evaluation word group assigned to one side view eye simulation image when one side view eye simulation image is evaluated in comparison with the other side view eye simulation image) based on the correlation (positive correlation, negative correlation, etc.) between the difference in unbiased variance V(θ) shown in the reference data (second reference data) and the score for each impression evaluation item included in the impression evaluation word group assigned to one side view eye simulation image).
For example, a pair of a first side view eye simulation image and a second side view eye simulation image are prepared, and V(θ) = 10 is calculated for the first side view eye simulation image and V(θ) = 110 for the second side view eye simulation image. If the difference in V(θ) is 20, 60, 140, or 200 in the reference data (second reference data), and the scores for the "intelligent" item are assigned -2, -1, +1, or +2, respectively, the score for the "intelligent" item in comparison between the first side view eye simulation image and the second side view eye simulation image can be calculated as 0.

In the counseling method of this embodiment, it is preferable to determine the content of suggestions regarding eyelash cosmetics and/or eyelash makeup methods based on the results of the impression evaluation.

The counseling method of this embodiment includes applying eyelash cosmetics and/or performing an eyelash makeup method on a subject to obtain a side view image of the eye makeup.
The side eye makeup image may be obtained by taking a new photograph of the subject wearing makeup, or in some cases, by applying image correction to the image obtained by taking a new photograph.
Here, the eyelash cosmetics and/or eyelash makeup methods applied to the subject may or may not comply with the content of the above proposals.

The counseling method of this embodiment includes performing an impression evaluation on a side view image of the eye makeup.

The impression evaluation of the side-view eye makeup image here may be the same as the impression evaluation embodiment in the impression evaluation method of this embodiment (see above), except for the embodiment specific to the counseling method of this embodiment.

In the impression evaluation in the counseling method of the present embodiment, a plurality of side-view eye makeup images having different unbiased variances V(θ) and/or different curl values S/Le are prepared,
The panelists evaluate the plurality of side eye makeup images with respect to each impression evaluation item included in the group of impression evaluation words,
creating reference data (first reference data) that indicates a correlation between the unbiased variance V(θ) and/or the curl value S/Le and an evaluation regarding each of the impression evaluation items included in the group of impression evaluation words;
It is preferable to use the reference data to perform an impression evaluation on the subject after makeup application.

In the impression evaluation in the counseling method of this embodiment, it is preferable to perform an absolute evaluation of one or more side-view eye makeup images with respect to each impression evaluation item included in the group of impression evaluation words based on reference data (first reference data).
Absolute evaluation may be performed, for example, by calculating the unbiased variance V(θ) for one or more side-view eye makeup images to be evaluated, and then determining the score for each impression evaluation item to be assigned to the calculated unbiased variance V(θ) in the reference data (first reference data).
For example, if V(θ) = 100 is calculated for the first side eye makeup image, and V(θ) = 20, 60, 140, or 200 in the reference data (first reference data), and the scores for the "intelligent" item are assigned as -2, -1, +1, or +2, respectively, the score for the "intelligent" item for the first side eye makeup image can be calculated as 0.

Furthermore, in this embodiment, it is preferable to classify the impression of the subject wearing makeup into types based on absolute evaluation.
Regarding the type used for classification, for example, an impression that is evaluated with a specific range of scores for the first impression evaluation item, a specific range of scores for the second impression evaluation item, a specific range of scores for the third impression evaluation item, and a specific range of scores for the fourth impression evaluation item can be set as a specific type.
For example, an impression that is evaluated as +1 to +3 for the "intelligent" item, +1 to +3 for the "decisive" item, +1 to +3 for the "refined" item, and +1 to +3 for the "clean" item can be considered to be an elegant type.

In the impression evaluation in the counseling method of the present embodiment, a plurality of side-view eye makeup images having different unbiased variances V(θ) and/or different curl values S/Le are prepared,
Regarding any pair of side eye makeup images among the plurality of side eye makeup images,
having panelists compare and evaluate one of the side view eye makeup images with the other side view eye makeup image with respect to each of the impression evaluation items included in the group of impression evaluation words;
creating reference data (second reference data) that indicates a correlation between a difference between the unbiased variance V(θ) and/or the curl value S/Le in the one side eye makeup image and the unbiased variance V(θ) and/or the curl value S/Le in the other side eye makeup image, and a comparative evaluation of each of the impression evaluation items included in the group of impression evaluation words;
It is preferable to use the reference data to perform an impression evaluation on the subject after makeup application.

In the impression evaluation in the counseling method of this embodiment, it is preferable to perform a comparative evaluation of multiple side-view eye makeup images with respect to each impression evaluation item included in the group of impression evaluation words based on reference data (second reference data).
The comparative evaluation may be performed, for example, by calculating the unbiased variance V(θ) of any pair of side eye images among the multiple side eye makeup images to be evaluated, then calculating the difference between the unbiased variances V(θ) of the two, and determining the score for each impression evaluation item of one of the pair of side eye makeup images in comparison with the other side eye makeup image (the score for each impression evaluation item included in the impression evaluation word group assigned to one side eye makeup image when one side eye makeup image is evaluated in comparison with the other side eye makeup image) based on the correlation (positive correlation, negative correlation, etc.) between the difference in unbiased variance V(θ) shown in the reference data (second reference data) and the score for each impression evaluation item included in the impression evaluation word group assigned to one side eye makeup image).
For example, a pair of a first side eye makeup image and a second side eye makeup image is prepared, and V(θ) is calculated as 10 for the first side eye makeup image and V(θ) as 110 for the second side eye makeup image. If the difference in V(θ) is 20, 60, 140, or 200 in the reference data (second reference data), and the scores for the "intelligent" item are assigned as -2, -1, +1, or +2, respectively, the score for the "intelligent" item in comparison between the first side eye makeup image and the second side eye makeup image can be calculated as 0.

In relation to the fact that the counseling method of this embodiment is carried out taking into consideration the unbiased variance V(θ) and/or curl value S/Le, it is preferable that the counseling method involves assessing the impression based on the unbiased variance V(θ) and/or curl value S/Le and suggesting an appropriate formulation.

The counseling method of this embodiment may use the aforementioned method for evaluating the degree of eyelash variation of this embodiment, the aforementioned method for measuring the degree of eyelash curl of this embodiment, and the aforementioned method for evaluating the impression of this embodiment.

(Market research methods)
The market research method of this embodiment includes:
In the side view image of the subject's eyes, the unbiased variance V(θ) of the angle θ between the horizontal line HL and the line segment BH connecting the base B and tip H of the eyelashes calculated for multiple eyelashes,
and/or
In the side view of the subject's eyes, the area S of the region surrounded by the line segment BH connecting the base B and the tip H of the eyelashes and the line tracing from the base B to the tip H of the eyelashes is divided by the length Le of the line tracing from the base to the tip of the eyelashes to obtain the curl value S/Le.
Prepare a plurality of side view images of the eye with different
This method investigates the correlation between the unbiased variance V(θ) and/or curl value S/Le and the evaluation of each impression evaluation item included in the group of impression evaluation words by having panelists evaluate multiple side-view images of the eyes with respect to each impression evaluation item included in the group of impression evaluation words.

Specifically, in this embodiment, it is preferable to prepare a plurality of side view eye images having different unbiased variances V(θ) and/or curl values S/Le, and have panelists evaluate the plurality of side view eye images with respect to each impression evaluation item included in the group of impression evaluation words, thereby investigating the correlation between the unbiased variances V(θ) and/or curl values S/Le obtained and the evaluations with respect to each impression evaluation item included in the group of impression evaluation words.
Here, when preparing multiple side view eye images having different unbiased variances V(θ) and/or curl values S/Le, side view eye images of multiple subjects having different unbiased variances V(θ) and/or curl values S/Le may be prepared, or multiple side view eye images having different unbiased variances V(θ) and/or curl values S/Le may be prepared by performing image correction on the side view eye image of one subject having a specific unbiased variance V(θ) and/or curl value S/Le.
Image correction may be performed using commercially available or free image analysis software.

The impression evaluation items included in the group of impression evaluation words are not particularly limited and may be selected appropriately depending on the purpose and use, but examples include intelligent, beautiful, motivated, proactive, decisive, gentle, honest, cheerful, friendly, nice, strong-willed, dignified, aesthetically pleasing, sophisticated, confident, approachable, clean, sociable, curious, and calm.
These impression evaluation items may be used alone or in combination of two or more to form the group of impression evaluation words.
The evaluation of each impression evaluation item may be performed qualitatively or quantitatively. When the evaluation is performed quantitatively, the evaluation may be performed by assigning a score (for example, from −3 to +3).

The survey method of this embodiment may investigate the correlation between the unbiased variance V(θ) and the evaluation regarding each impression evaluation word included in the impression evaluation word group, may investigate the correlation between the curl value S/Le and the evaluation regarding each impression evaluation word included in the impression evaluation word group, or may investigate the correlation between the unbiased variance V(θ) and the curl value S/Le and the evaluation regarding each impression evaluation word included in the impression evaluation word group.
The investigation method of this embodiment is preferably carried out using two or more side view images of the eyes, more preferably four or more side view images of the eyes, and even more preferably six or more side view images of the eyes.
Here, the correlation in the research method of this embodiment is not particularly limited, and may be a positive correlation, a negative correlation, or a relationship that provides reasonable predictability. The strength of the correlation is also not particularly limited.

In addition, in the market research method of this embodiment, the unbiased variance V(θ) of the angle θ between the horizontal line HL and the line segment BH connecting the base B and tip H of the eyelashes calculated for a plurality of eyelashes in the side view image of the subject's eyes,
and/or
In the side view of the subject's eyes, the area S of the region surrounded by the line segment BH connecting the base B and the tip H of the eyelashes and the line tracing the line from the base B to the tip H of the eyelashes is divided by the length Le of the line tracing the line from the base to the tip of the eyelashes to obtain the curl value S/Le.
Prepare a plurality of side view images of the eye with different
For any pair of the side eye images among the plurality of side eye images,
having the panelists compare and evaluate one of the side view images of the eyes with the other side view image of the eyes with respect to each of the impression evaluation items included in the group of impression evaluation words;
It is preferable to investigate the correlation between the difference between the unbiased variance V(θ) and/or the curl value S/Le in one of the side eye images and the unbiased variance V(θ) and/or the curl value S/Le in the other side eye image and the comparative evaluation of each of the impression evaluation items included in the group of impression evaluation words.

In the market research method of this embodiment, the unbiased variance V(θ) of the angle θ between the horizontal line HL and the line segment BH connecting the base B and tip H of the eyelashes calculated for a plurality of eyelashes in a side view image of the subject's eyes,
and/or
In the side view of the subject's eyes, the area S of the region surrounded by the line segment BH connecting the base B and the tip H of the eyelashes and the line tracing the line from the base B to the tip H of the eyelashes is divided by the length Le of the line tracing the line from the base to the tip of the eyelashes to obtain the curl value S/Le.
A plurality of side eye images with different
an average side view eye image obtained by averaging the plurality of side view eye images having different unbiased variances V(θ) and/or different curl values S/Le;
Prepare
It is preferable to have the panelists compare each of the side eye images with the average side eye image, thereby having the panelists evaluate the plurality of side eye images with respect to each impression evaluation item included in the group of impression evaluation words.

In the market research method of this embodiment, it is preferable that the panelists be at least one selected from the group consisting of experts, researchers, beauty consultants, and consumers.

The above describes exemplary embodiments of the method for evaluating the degree of eyelash variation, the method for evaluating the degree of eyelash curl, the impression evaluation method, the counseling method, and the market research method of the present invention, with reference to the drawings. However, the above embodiments can be modified as appropriate, and the present invention is not limited to the above exemplary embodiments.

The present invention will be explained in more detail below using examples, but the present invention is not limited to the following examples in any way.

(Acquisition of eye images)
In 2019, an in-house skin quality survey was conducted and images of the eyes of 65 subjects (women, aged 20 to 50) were obtained.
The subject was asked to remove their makeup, and their face was fixed to a face stand that could fix the chin and forehead. A jig with a length correction sticker attached was placed in a predetermined position relative to the fixed face.
With the eyes open, photographs were taken from the side and front using the camera function of an iPod (registered trademark) as described below.

((Side view of the eyes))
As shown in Figure 3(A), the above-mentioned jig for side photography was placed so that the length correction sticker was positioned at the same height as the center of the pupil on the side of the face, so that the length correction sticker overlapped the center of the pupil when viewed from the front of the face, and so that the surface of the length correction sticker faced the side of the face.
The camera was positioned so that the lens could capture the subject's eyes and the length correction stickers.
The distance between the top of the pupil (the outermost edge of the pupil at the front of the face) and the length correction sticker was 5 cm.
The distance between the top of the pupil (the outermost front edge of the pupil on the face) and the camera lens was 10 cm.
Under the above conditions, a photograph of the subject's eyes was taken from the side, taken from a direction along the line connecting the tops of the subject's left and right ears.

((Front view of eyes))
As shown in Figure 3(B), the jig for frontal photography was set so that the length correction sticker was positioned at the same height as the center of the pupil on the side of the face, so that the length correction sticker overlapped with the center of the pupil when viewed from the side of the face, and so that the surface of the length correction sticker faced the front of the face.
The camera was positioned so that the lens could capture the subject's eyes and the length correction stickers.
The distance between the top of the pupil (the outermost edge of the pupil at the front of the face) and the length correction sticker was 5 cm.
The distance between the top of the pupil (the outermost front edge of the pupil on the face) and the camera lens was 10 cm.
Under the above conditions, a photograph of the subject's eyes was taken from the front, in a direction perpendicular to the line connecting the tops of the left and right ears and along the horizontal line.

The side and front images of the eyes obtained by photography were subjected to image analysis using Image J as follows.
First, length correction was performed on all side and front eye images based on the dimensions of the length correction sticker.
In the side view of the eye, the area within 2 mm (see dashed circle in Figure 2(A)) from the outermost horizontal edge of the upper eyelid (see "X" mark in Figure 2(A)) was identified as the apex of the eye.
Ten eyelashes with the bases B located at the top of the eye in the side view of the eye were selected based on the length of the line segment BH from the first to tenth longest lengths, and used to measure the unbiased variance V(θ), which will be described later. Furthermore, of these ten eyelashes, the eyelashes with the length of the line segment BH from the first to third longest lengths were selected and used to measure the curl value S/Le, which will be described later.
In addition, the area within 2 mm (see dashed circle in Figure 2(B)) from the midpoint of the line tracing the upper eyelid from its beginning (inner corner of the eye) to its end (outer corner of the eye) in the frontal eye image was identified as the central part of the upper eyelid.
As described above, the 10 eyelashes selected in the side view eye image were also located in the center of the upper eyelid in the front view eye image.

(Measurement of unbiased variance V(θ))
For the ten eyelashes located at the top of the eye and in the center of the upper eyelid, the unbiased variance V(θ) of the angle θ formed by the line segment BH connecting the base B and tip H of the eyelash and the horizontal line HL was calculated using the following formula (1).

...(1)

In equation (1), θi represents the angle θ for the i-th eyelash, and AVE(θ) represents the average angle θ for n eyelashes. Here, n = 10.

The larger the unbiased variance V(θ), the greater the degree of variation in the eyelashes of the subjects was evaluated.
The results are shown in Table 1.

At the same time, visual evaluation was also conducted by expert panelists, and the evaluation was based on the following criteria: "Good": there was variation in the eyelashes of the subjects; "Poor": there was no variation in the eyelashes of the subjects.
The results are shown in Table 1.

While visual evaluation by expert panelists makes it difficult to determine the boundary between large and small degrees of eyelash variation, the evaluation method of the present invention has been shown to make it possible to continuously determine the degree of eyelash variation.

(Measurement of curl value S/Le)
For the three eyelashes located at the top of the eye and in the center of the upper eyelid, the area S of the region surrounded by the line segment BH connecting the base B and tip H of the eyelashes and the line tracing from the base B to the tip H of the eyelashes was calculated by determining the length and area through image analysis using ImageJ.
Furthermore, for each of the three eyelashes, the length Le of the line tracing the eyelash from the base B to the tip H was calculated by image analysis using ImageJ.
The calculated area S was divided by the calculated Le to calculate the curl value S/Le.

The larger the curl value S/Le, the greater the degree of curl of the eyelashes was evaluated.
The results are shown in Table 2.

At the same time, visual evaluation was also conducted by expert panelists, and the evaluation was based on the following criteria: "Score 3": the subject's eyelashes were significantly curled; "Score 2": the subject's eyelashes were slightly curled; and "Score 1": the subject's eyelashes were not curled.
The results are shown in Table 1.

While visual evaluation by expert panelists makes it difficult to judge the difference of one score in the degree of eyelash curl, the evaluation method of the present invention has been shown to make it possible to continuously judge the degree of eyelash curl.

(Creating standard data for impression evaluation methods, market research)
Using a side view image of the eye of a subject (female, in her 20s), the degree of eyelash variation was changed by image modification to create side view simulation images of the eye (subject C-No. 1, subject C-No. 2, subject C-No. 3) with unbiased variances V(θ) different from those before the change, and the unbiased variance V(θ) was calculated for each side view simulation image of the eye.
Using a side view image of the eye of a subject (female, in her 20s), the degree of eyelash variation was changed by image modification to create side view simulation images of the eye (subject C-No. 4, subject C-No. 5, subject C-No. 6) with unbiased variances V(θ) different from those before the change, and the unbiased variance V(θ) was calculated for each side view simulation image of the eye.
Then, for all the side-view eye simulation images, 31 expert panelists visually evaluated each impression evaluation item included in a group of impression evaluation words prepared in advance (intelligent, beautiful, motivated, proactive, decisive, gentle, honest, cheerful, friendly, pleasant, strong-willed, dignified, aesthetically pleasing, refined, confident, approachable, clean, sociable, curious, and calm).
In the visual evaluation, each impression evaluation word was evaluated by giving a score of −3 to +3.
The results are shown in Table 3.

According to the present invention, it is possible to provide a method for accurately and simply evaluating the degree of curl and the degree of variation of eyelashes.
The present invention also provides a method for evaluating an impression of a subject based on the degree of eyelash curl and the degree of eyelash variation. Furthermore, the present invention also provides a method for providing counseling to a subject based on the degree of eyelash curl and the degree of eyelash variation. Furthermore, the present invention also makes it possible to investigate the correlation between the degree of eyelash curl and the degree of eyelash variation and the evaluation of each impression evaluation item included in a group of impression evaluation words.

B: Base of eyelashes H: Tip of eyelashes BH: Line segment HL: Horizontal line θ: Angle Le: Length of the line tracing from base B of eyelashes to tip H S: Area of the region enclosed by line segment BH and the line tracing from base B of eyelashes to tip H

Claims (29)

An image analysis method using an information processing device,
calculating an unbiased variance V(θ) of an angle θ formed by a line segment BH connecting the base B and tip H of each of a plurality of eyelashes with a horizontal line HL in a side view image of the subject's eye;
and outputting information corresponding to a degree of variation in the eyelashes of the subject based on the unbiased variance V(θ).
Image analysis methods. The image analysis method according to claim 1, wherein the plurality of eyelashes are eyelashes whose bases B are located at the apex of the eye. The image analysis method according to claim 1 or 2, wherein the angle θ of all of the multiple eyelashes is in the range of -40° to +40° (horizon is 0°), and when all of the eyelashes are ranked from smallest to largest angle θ and the difference in angle θ between the ranked positions is calculated as an absolute value, the average difference is 10° or less. The image analysis method described in any one of claims 1 to 3, wherein the plurality of eyelashes are eyelashes located in the center of the upper eyelid in the frontal eye image of the subject. An image analysis method using an information processing device,
calculating a curl value S/Le by dividing an area S of a region surrounded by a line segment BH connecting the base B and tip H of one or more eyelashes and a line tracing the line from the base B to the tip H of the eyelash in the side view image of the eye of the subject, by a length Le of the line tracing the line from the base B to the tip H of the eyelash;
and outputting information corresponding to the degree of curl of the eyelashes based on the curl value S/Le.
Image analysis methods. The image analysis method according to claim 5, wherein information corresponding to the degree of curl of the eyelashes is output based on the average value of the curl values S/Le calculated for the plurality of eyelashes. The image analysis method described in claim 5 or 6, wherein the plurality of eyelashes are eyelashes whose bases B are located at the top of the eye in the side view eye image. The image analysis method of claim 7, wherein the plurality of eyelashes are those having the first to third longest lengths of the line segment BH among the eyelashes located at the top of the eye, and the angle θ for all of the eyelashes is in the range of -40° to +40° (horizon is 0°), and when all of the eyelashes are ranked from smallest to largest angle θ and the difference in angle θ between the ranked positions is calculated as an absolute value, the average difference is 10° or less. The image analysis method described in claim 8, wherein the plurality of eyelashes are eyelashes located in the center of the upper eyelid in the frontal eye image of the subject. An image analysis method using an information device,
In the side view image of the subject's eye, the unbiased variance V(θ) of the angle θ between the horizontal line HL and the line segment BH connecting the base B and tip H of the eyelashes calculated for multiple eyelashes.
and/or
In the side view of the eye of the subject, the area S of the region surrounded by the line segment BH connecting the base B and the tip H of the eyelash and the line tracing the line from the base B to the tip H of the eyelash is divided by the length Le of the line tracing the line from the base B to the tip H of the eyelash, to obtain a curl value S/Le;
outputting information corresponding to the subject's impression based on the
Image analysis methods. The plurality of side view eye images have different unbiased variances V(θ) and/or different curl values S/Le,
acquiring evaluation information regarding each impression evaluation item included in a group of impression evaluation words of the plurality of side eye images;
creating reference data indicating a correlation between the unbiased variance V(θ) and/or the curl value S/Le and an evaluation regarding each of the impression evaluation items included in the group of impression evaluation words;
The image analysis method according to claim 10 , further comprising the step of: outputting information corresponding to the subject's impression using the reference data. The image analysis method according to claim 11 , wherein the evaluation of the impressions of the plurality of side eye images is an absolute evaluation of each impression evaluation item included in the group of impression evaluation words based on the reference data. The image analysis method described in claim 12, wherein the absolute evaluation can be used to classify the subject's impression. The plurality of side view eye images have different unbiased variances V(θ) and/or different curl values S/Le,
For any pair of the side eye images among the plurality of side eye images,
acquiring information on a comparative evaluation of one of the side view eye images with respect to impression evaluation items by comparing it with the other side view eye image;
creating reference data showing a correlation between a difference between the unbiased variance V(θ) and/or the curl value S/Le in the one side view eye image and the unbiased variance V(θ) and/or the curl value S/Le in the other side view eye image, and a comparative evaluation of each impression evaluation item included in a group of impression evaluation words;
and outputting information corresponding to the subject's impression using the reference data.
The image analysis method according to claim 10. The image analysis method according to claim 14 , wherein the evaluation of the impressions of the plurality of side view eye images is a comparative evaluation for each impression evaluation item, performed based on the reference data. The image analysis method described in any one of claims 11 to 15, further comprising a step of updating the reference data periodically or irregularly. 11. The image analysis method according to claim 10 , further comprising a step of modifying the side view image of the eye of the subject by image correction to change the degree of eyelash variation and/or curl degree, thereby obtaining a side view simulation image of the eye having the unbiased variance V(θ) and/or the curl value S/Le different from those before the modification. the side eye simulation images have different unbiased variances V(θ) and/or different curl values S/Le,
outputting evaluation information regarding each impression evaluation item included in a group of impression evaluation words of the plurality of side eye simulation images;
creating reference data indicating a correlation between the unbiased variance V(θ) and/or the curl value S/Le and an evaluation regarding each of the impression evaluation items included in the group of impression evaluation words;
and outputting information corresponding to another impression of the subject using the reference data.
The image analysis method according to claim 17. the evaluation of the impressions of the plurality of side eye simulation images is an absolute evaluation for each impression evaluation item included in the group of impression evaluation words, performed based on the reference data;
The image analysis method according to claim 18. the side eye simulation images have different unbiased variances V(θ) and/or different curl values S/Le,
For any pair of the side view eye simulation images among the plurality of side view eye simulation images,
acquiring information on a comparative evaluation of each impression evaluation item included in a group of impression evaluation words of one of the side view eye simulation images by comparing it with the other side view eye simulation image;
creating reference data indicating a correlation between a difference between the unbiased variance V(θ) and/or the curl value S/Le in one of the side view eye simulation images and the unbiased variance V(θ) and/or the curl value S/Le in the other side view eye simulation image, and a comparative evaluation of each of the impression evaluation items included in the group of impression evaluation words;
and outputting information corresponding to the subject's impression after the simulation using the reference data.
The image analysis method according to claim 17. The image analysis method according to claim 20, wherein the evaluation of the impressions of the plurality of side eye simulation images is a comparative evaluation of each impression evaluation item included in the group of impression evaluation words based on the reference data. The image analysis method according to claim 10 , further comprising a side eye makeup image in which eyelash cosmetics and/or an eyelash makeup method have been applied to the side eye image of the subject. The image analysis method described in claim 22, wherein the impression of the side view eye image includes the impression of the side view eye makeup image. the plurality of side eye makeup images have different unbiased variances V(θ) and/or different curl values S/Le,
acquiring evaluation information regarding each impression evaluation item included in a group of impression evaluation words of the plurality of side eye makeup images;
creating reference data indicating a correlation between the unbiased variance V(θ) and/or the curl value S/Le and an evaluation regarding each of the impression evaluation items included in the group of impression evaluation words;
and outputting information corresponding to an impression of the subject after makeup application using the reference data.
24. The image analysis method according to claim 23. The image analysis method described in claim 24, wherein the impression of the side eye makeup image is an absolute evaluation of each impression evaluation item included in the group of impression evaluation words based on the reference data. the plurality of side eye makeup images have different unbiased variances V(θ) and/or different curl values S/Le,
For any pair of side eye makeup images among the plurality of side eye makeup images,
a step of acquiring information on a comparative evaluation of one of the side eye makeup images with respect to each impression evaluation item included in a group of impression evaluation words, with respect to the other side eye makeup image;
creating reference data indicating a correlation between a difference between the unbiased variance V(θ) and/or the curl value S/Le in the one side eye makeup image and the unbiased variance V(θ) and/or the curl value S/Le in the other side eye makeup image, and a comparative evaluation of each of the impression evaluation items included in the group of impression evaluation words;
23. The image analysis method according to claim 22, further comprising: using the reference data to output information corresponding to an impression of the subject after the simulation. An image analysis method using an information processing device,
In the side view image of the subject's eyes, the unbiased variance V(θ) of the angle θ between the horizontal line HL and the line segment BH connecting the base B and tip H of the eyelashes calculated for multiple eyelashes,
and/or
acquiring a plurality of side view images of the subject's eyes with different curl values S/Le, which is calculated by dividing an area S of a region surrounded by a line segment BH connecting the base B and the tip H of the eyelashes and a line tracing the line from the base B to the tip H of the eyelashes by a length Le of the line tracing the line from the base to the tip of the eyelashes;
a step of outputting information corresponding to a correlation between the unbiased variance V(θ) and/or the curl value S/Le and an evaluation of each of the impression evaluation items included in the group of impression evaluation words, based on an evaluation of the plurality of side view eye images with respect to each of the impression evaluation items included in the group of impression evaluation words;
An image analysis method comprising: An image analysis method using an information processing device,
In the side view image of the subject's eyes, the unbiased variance V(θ) of the angle θ between the horizontal line HL and the line segment BH connecting the base B and tip H of the eyelashes calculated for multiple eyelashes,
and/or
In the side view of the subject's eyes, the area S of the region surrounded by the line segment BH connecting the base B and the tip H of the eyelashes and the line tracing the line from the base B to the tip H of the eyelashes is divided by the length Le of the line tracing the line from the base to the tip of the eyelashes to obtain the curl value S/Le.
acquiring a plurality of side eye images with different
For any pair of the side eye images among the plurality of side eye images,
acquiring information on a comparative evaluation of one of the side view images of the eyes with respect to each impression evaluation item included in a group of impression evaluation words;
and outputting information corresponding to a correlation between a difference between the unbiased variance V(θ) and/or the curl value S/Le in the one side view eye image and the unbiased variance V(θ) and/or the curl value S/Le in the other side view eye image, and a comparative evaluation of each impression evaluation item included in the impression evaluation word group.
Image analysis methods. An image analysis method using an information processing device,
In the side view image of the subject's eyes, the unbiased variance V(θ) of the angle θ between the horizontal line HL and the line segment BH connecting the base B and tip H of the eyelashes calculated for multiple eyelashes,
and/or
In the side view of the subject's eyes, the area S of the region surrounded by the line segment BH connecting the base B and the tip H of the eyelashes and the line tracing the line from the base B to the tip H of the eyelashes is divided by the length Le of the line tracing the line from the base to the tip of the eyelashes to obtain the curl value S/Le.
A plurality of side eye images with different
an average side view eye image obtained by averaging the plurality of side view eye images having different unbiased variances V(θ) and/or different curl values S/Le;
and obtaining
outputting information corresponding to an evaluation made for each impression evaluation item included in the group of impression evaluation words of the plurality of side eye images based on a comparison between each side eye image and the average side eye image;
An image analysis method comprising: