TWI471807B - Face classification method, face classification device, classification chart, face classification program and recorded the program of the recording media - Google Patents

Description

Face classification method, face classification device, classification map, face classification program, and recording medium on which the program is recorded Technical field

The present invention relates to a face classification method, a face classification device, a classification map, a face classification program, and a recording medium on which the program is recorded, and in particular, a face classification method for classifying a face with a high degree of precision. A face classification device, a classification map, a face classification program, and a recording medium on which the program is recorded.

Background technique

In the past, it has been mainly used to emphasize the three-dimensionality by using light or shadow, as a makeup technique for expressing the three-dimensionality of the face or the sensitivity of the flesh. Specifically, there is a method of making the nose appear light and shadow, causing the nose to appear a shadow to improve the visual effect of the nose, causing the cheek to appear a shadow to present a sharp facial line, and applying an eye shadow to the raised eyelid to reveal the eyelid depression effect. A method of presenting a depth profile, and the like.

In addition, there is a method of classifying the face of the subject by the feature of the face type, and displaying the guiding principle of the makeup according to the classification type (see, for example, Patent Document 1 and Patent Document 2).

Patent Document 1: Japanese Patent Laid-Open Publication No. 2001-224430

Patent Document 2: Japanese Patent Laid-Open Publication No. 10-289303

However, according to the aforementioned makeup technique which expresses the three-dimensional or shadowy feeling of the face, it may be caused by excessive emphasis on the three-dimensional feeling and not on the overall balance. Good result. Further, the above-described makeup technique is a technique for a part of a face such as a nose or an eye, and is not a technique of capturing a three-dimensional feeling of the entire face.

Therefore, it is necessary to establish an index which can accurately capture the stereoscopic characteristics of various entire faces and a method of adjusting the stereoscopic feeling of the entire face. However, although the classification map is used as one index as described above, the classification map corresponding to the stereoscopic effect of the entire face is not disclosed.

The present invention has been made in view of the above problems, and an object thereof is to provide a face classification method, a face classification device, a classification map, a face classification program, and a program for recording a face that can perform high-precision classification of facial stereoscopic effects. Record media.

In order to solve the aforementioned problems, the present invention employs a method for solving the problems having the following features.

One embodiment of the present invention is a method for classifying a face that is classified into a three-dimensional sense of a subject's face, and includes: an evaluation step, which is based on at least a depth of the face space and the entire face, and a sensitivity of the flesh and the skin. The index is used to evaluate the face of the subject; and the classification step is to locate the result of the evaluation by the foregoing evaluation step, and to locate at least one of the at least four types that have been classified in advance.

According to this embodiment, the three-dimensionality of the face can be classified with high precision based on at least two indicators of the depth of the face space and the depth of the face and the sensitivity of the skin.

In another embodiment of the present invention, the evaluation step uses the average face set in advance and each of the four types of the aforementioned types, The face of the subject was evaluated.

According to the present embodiment, the evaluation accuracy can be improved and the evaluation can be performed simply and efficiently by using the average facial image and each of the four types of representative facial images.

In another embodiment of the present invention, the evaluation step may reflect the ratio of the length of the face width of the balance of the forehead, cheek or chin of the face space according to the average face evaluation. The depth and the degree of softness are evaluated by the degree of sharpness or the degree of softness.

According to this embodiment, the stereoscopic feeling can be evaluated by an appropriate scale.

Furthermore, in another embodiment of the present invention, a display step of displaying an adjustment method for adjusting each of the four types of stereoscopic feelings by makeup is further included, and the face type classified by the classification step is correspondingly displayed on Each of the quadrants of the plane coordinate space has been positioned with the aforementioned average face and the above four types of drawings, and the three-dimensional sense of each of the above four types is closer to the stereoscopic feeling of the average face, and is displayed with makeup. The method of adjusting the stereoscopic feeling of each of the above four types is adjusted.

According to this embodiment, the makeup method can be easily grasped by displaying the adjustment method of adjusting the three-dimensional feeling by makeup.

In another embodiment of the present invention, the adjustment method is adjusted according to a shape similar to the contour shape of the average face, and the width reduced by a predetermined ratio is formed to have a length ratio of 1:1.35 to 1.45.

According to this embodiment, the contour of the face can be adjusted to the inner side of the face by adjusting the visual effect of the face space and the depth and the sensitivity of the flesh, and an appropriate three-dimensional feeling can be obtained.

Further, another embodiment of the present invention is directed to a face classification device for classifying a stereoscopic feeling of a subject's face, comprising: an imaging mechanism capable of capturing a face image of the subject The control mechanism is capable of positioning the image of the face after shooting, at least according to the depth of the face space and the depth of the face as a whole, and the sensitivity of the flesh and blood, to control at least one of the at least four types that have been classified in advance. And the display mechanism, wherein the classification results classified by the foregoing control mechanism are correspondingly displayed in the quadrants of the plane coordinate space, and the above four types of maps are positioned.

According to this embodiment, the three-dimensional sense of the face can be classified with high precision according to at least two indexes of the face space and the depth of the whole face and the sensitivity of the flesh.

In another embodiment of the present invention, the control means uses the average face set in advance and each of the four types of representative faces to evaluate the face of the subject.

According to the present embodiment, the evaluation accuracy can be improved and the evaluation can be performed simply and efficiently by using the average facial image and each of the four types of representative facial images.

In another embodiment of the present invention, the control mechanism evaluates a ratio of a face width that can reflect the length and width of the forehead, cheek, or chin of the face space based on the average face, and The depth and the flesh sensitivity are evaluated by the degree of sharpness or the degree of softness.

According to this embodiment, the stereoscopic feeling can be evaluated by an appropriate scale.

In another embodiment of the present invention, the display mechanism is based on the classified face type, in order to be closer to the stereoscopic shape of the average face. The visual effect of the sense is displayed, and the method of adjusting the stereoscopic feeling of each of the above four types is adjusted by makeup.

According to this embodiment, the makeup method can be easily grasped by displaying the adjustment method of adjusting the three-dimensional feeling by makeup.

In another embodiment of the present invention, the adjustment method is adjusted according to a shape similar to the contour of the average face, and the width reduced by a predetermined ratio is formed to have a length ratio of 1:1.35 to 1.45.

According to the present embodiment, the contour of the face can be adjusted by adjusting the visual effect of the face space and the depth and the sensitivity of the flesh, and an appropriate three-dimensional feeling can be obtained.

Another embodiment of the present invention includes: a first axis that displays a face type including a face space; and a second axis that displays a face type formed by the depth of the entire face and the sensitivity of the flesh; And the coordinate space is a person who has a face image, and the face image is set in advance in the quadrant of the two axes and the center of the two-axis intersection, and has a standard and ideal three-dimensional feeling.

According to this embodiment, the three-dimensional sense of the face can be classified with high precision according to the depth of the face space and the depth of the whole face and the sensitivity of the flesh and the skin.

In another embodiment of the present invention, the first axis system is configured to reflect a face width of the forehead, the cheek or the chin as the front face of the face space, which is balanced with respect to a predetermined face. The ratio of the size is small; while the second axis shows the depth and the sharpness of the flesh sensitivity, and the other side shows the fullness and softness.

According to this embodiment, the stereoscopic feeling can be evaluated by an appropriate scale.

In another embodiment of the present invention, the method of adjusting the stereoscopic effect by the makeup can be displayed on the quadrant classified by the two axes to approximate the stereoscopic feeling of the average face.

According to the present embodiment, the visual effect of the face space, the depth, and the sensitivity of the flesh can be adjusted according to the figure, and the makeup with good stereoscopic effect can be performed.

Further, another embodiment of the present invention enables a computer to operate the face classification program of the face classification method of the present invention.

According to this embodiment, the three-dimensionality of the face can be classified with high precision. Further, by installing the program, the face portion of the present invention can be easily realized by a general-purpose personal computer or the like.

Further, another embodiment of the present invention is a recording medium on which a face classification program of the present invention is recorded and which can be read by a computer.

According to this embodiment, the face classification program can be easily installed to most other computers by the recording medium. Further, by installing the program, the face portion of the present invention can be easily realized by a general-purpose personal computer or the like.

According to the present invention, the stereoscopic feeling of the face can be classified with high precision.

Simple description of the schema

Fig. 1 is a view showing an example of the device configuration of the face classification device of the present invention.

Fig. 2A shows an example of a feature expression term (1) that is recognized by a facial image.

Fig. 2B shows an example (2) of the feature expressions that are recognized by the face image.

Fig. 3 is a view showing an example of an evaluation table for evaluation of the three-dimensionality of the face.

Fig. 4 is a view showing an example of the average evaluation result of the object.

Fig. 5 is a view showing an example of evaluation results of physical objects and photographs.

Fig. 6 is a view showing an example of a stereoscopic evaluation table based on the selected language.

Figure 7 shows the factor load.

Fig. 8 shows an example in which each object is drawn by the number of scores of the first and second factors obtained.

Fig. 9 shows an example in which each object is drawn by the number of scores of the third and fourth factors obtained.

Fig. 10 is a view showing a first embodiment of a classification map classified according to the depth of the face, the sensitivity of the flesh and the space.

The eleventh (a) to (d) diagrams show an example of the focus of each index of the present invention at the time of evaluation.

Fig. 12 is an example of an evaluation item showing a visual effect and an overall impression on the face.

Fig. 13 shows an example of the evaluation result regarding the visual effect of the face.

Fig. 14 is an example of an "average face" made using a face photograph of 40 women.

Figure 15 shows the information on the depth and the sensitivity of the flesh and blood from the "average face".

Fig. 16 is a view showing an example of the result of the shadow analysis of the average face.

Figure 17 shows an example of the best elliptical shape.

Fig. 18 is a view for explaining an example of the optimum elliptical shape.

Fig. 19 is an example of a record showing the application method of facial powder.

Figure 20 shows the appearance of the person A and the application of facial powder after evaluation. An example of the face.

Fig. 21 is a view showing a second embodiment of a classification map classified according to the depth of the face, the sensitivity of the flesh and the space.

The 22nd (a) to (e) figure shows an example of the makeup adjustment method for each type of face part.

The 23rd (a) to (b) figure shows an example of the evaluation of the subject of the cosmetic product.

Figure 24 is a flow chart showing the makeup sequence using the classification chart of the present invention.

The best form of implementing the invention

Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings.

<Outline of the Invention>

The present invention focuses on the point of visual effect of adjusting the three-dimensionality of the face by makeup, and evaluates the stereoscopic feeling of each face, and classifies the classification map by setting the result in advance, and can be classified according to the classification type. Simple and efficient adjustment of the stereoscopic effect.

Here, the facial evaluation and classification index of the present invention are expressed in terms of facial space and depth, and the sensitivity of the flesh and blood, specifically, according to "the length of the face space is long or short", and the depth and the sensitivity of the flesh are thin and sharp. Still full of softness and judgment. Here, the depth, the flesh sensitivity is a composite index of both "depth" and "bone sensitivity", and the two indicators are related. That is, it has a skinny (no flesh) and has a depth without a skinfeel (having a fleshy feeling) without a depth relationship, so the composite table is based on the correlation. Present.

Further, in the center of the quadrant classified by the two axes of the graph and the center of the two-axis intersection, an example of a face image which is set in advance and has an average face with a standard and ideal stereoscopic effect is arranged, and the evaluation accuracy is improved and is simple. Evaluate efficiently.

Further, on the classification map of the stereoscopic feeling, the method of adjusting the stereoscopic effect (the essentials) which is set in advance according to each different facial feature (type) by the makeup adjustment is recorded. Thereby, it is possible to easily grasp what kind of makeup is to be performed for each classification.

Furthermore, the adjustment of the stereoscopic effect uses the "best elliptical adjustment method". Here, the optimal ellipse represents a facial contour shape having a standard face balance. The contour shape is similar to the pattern of two super-ellipse composites having different curvatures on the upper and lower sides of the face, that is, the shape reduced by a predetermined ratio and the width-to-width ratio of 1:1.4. . In order to realize such an optimal elliptical shape, it is possible to adjust the visual effect of the visual effect of the facial space and the depth and the sensitivity of the flesh and the skin by a picture or paper, and to achieve a three-dimensional makeup.

<device construction>

First, a device for inputting the evaluation information on the face of the subject, and analyzing and evaluating the classification belonging to the map based on the classification map prepared in advance to further display the face classification device by the stereoscopic adjustment method of makeup Construction example. In addition, the face classification device to be described later can also be used when analyzing processing for producing a classification map to be described later.

Here, the face classification device uses a personal computer such as a general-purpose computer, a server, etc., and can be installed to enable the computer to perform various processes related to the present invention. The program (face classification program) is executed to implement the face classification processing of the present invention.

Fig. 1 is a view showing an example of the device configuration of the face classification device of the present invention. The face classification device 10 of Fig. 1 includes an input device 11, an output device 12, a drive device 13, an auxiliary memory device 14, a memory 15, a CPU 16 (central processing unit) for performing various controls, a network connection device 17, and a photographing device. 18, and the devices are connected to each other through the system bus bar B.

The input device 11 has a pointing device such as a keyboard and a mouse for the user to operate, and can input various operation signals such as program execution from the user. The output device 12 has a display for displaying various required windows or materials and the like when the computer body is operated to perform the processing of the present invention, and can display the execution progress or result of the program of the control program possessed by the CPU 16. Further, the output device 12 may have a function of printing or the like. In this case, the evaluation table, the face classification map, the face classification result, the adjustment method of the three-dimensional feeling by makeup, or the like can be printed on the various information that can be obtained. A print medium such as paper is provided to an experimentee or the like.

Here, the execution program installed in the computer body of the present invention is provided by a recording medium 19 such as a CD-ROM. The recording medium 19 on which the program is recorded can be installed in the drive device 13, and the execution program included in the recording medium 19 is mounted from the recording medium 19 to the auxiliary storage device 14 via the drive device 13. Further, the recording medium 19 may be a recording medium such as a floppy disk or a magnetic optical disk that can record information optically, electrically, or magnetically, such as a CD-ROM; ROM (read only memory); flash memory, etc. Various types of recording media such as semiconductor memories that electrically record information.

Moreover, the auxiliary memory device 14 is a storage mechanism such as a hard disk, and can store the book. The image data of the invention, the control program set on the computer, the average facial image, and the evaluation results of the face of the subject are input and output as needed.

The CPU 16 can control various operations such as various calculations or input/output of data of each hardware component according to a control program such as an OS (operation system) and an execution program read and stored by the memory 15 to realize a face. Various processing will be described later in the classification processing and the like. Further, the CPU 16 can acquire various kinds of information such as the face evaluation result of the subject required to execute the program from the auxiliary storage device 14, or the required information when the classification chart of the present invention is displayed on the output device 12, and the like. The execution result of the program or the like is stored in the auxiliary memory device 14.

The network connection device 17 can obtain an execution program from a network or the like, or other terminal connected to a communication network represented by a LAN (regional network) or the like by connecting to a communication network or the like, and can obtain an execution program by executing the program. The execution result or the execution program of the present invention itself is provided to other terminals and the like.

Further, the imaging device 18 is constituted by a CCD (Charge Coupled Device) camera or the like, and can capture a front face image of the subject or an inclined face image with respect to a predetermined angle. The image after the shooting is stored in the auxiliary memory device 14 or the like, and is read out as needed to perform face classification and the like.

The face classification processing of the present invention can be performed by the aforementioned device configuration. Further, by installing the program, the face classification processing of the present invention can be easily realized by using a general-purpose personal computer or the like.

<Face classification method of the present invention>

Next, a classification map for the face classification method of the present invention and a guideline for a method of adjusting a three-dimensional effect by makeup, and the present invention are included The contents of each review are also specified.

<1. Analysis of the three-dimensional facial expression and review of the classification method> <1-1. Sample collection for obtaining evaluation terms>

First, construct indicators for the face class. Collect facial photos of people with different facial features to review what kind of performance should be used to capture the overall three-dimensionality of the face. Specifically, 21 women aged 20 to 40 working in a cosmetic company collected face samples taken from the front and from a direction inclined at 45 degrees.

<1-2. Preparatory review of analysis indicators> <1-2-1. Understanding the common understanding of the three-dimensional sense>

The present invention evaluates and classifies each face according to the characteristics of the three-dimensionality of the face. Therefore, it is premised on having a common understanding in the sense of three-dimensionality. Therefore, under the respective benchmarks of the four research leaders, most of the facial photos that feel "similar in three-dimensional face" are grouped together, and more similar groups are placed in the vicinity to review whether they can be The result is recognized as commonality. In addition, the review used the front face photos of the 21 samples collected.

Based on the results of the four people, it is known that the faces of the same group are shared by the respective persons in charge, or the faces that are located at the close position even if the groups are different. As can be seen from the above, it has common visibility in the judgment of the three-dimensionality of the face.

<1-2-2. Presumption of evaluation index>

Next, for the similar features of the face focused on the group, the respective photos are classified and photographed and written on the printing paper. Result, Knowing the "feeling and bone sensitivity of the whole face", "the width and length of the face", "the shape of the face line", "the width of the cheek or forehead", etc. In addition, it is known that there is also a face that is judged to be a standard type. This review is used as a reference for the integration of the results of the review and later analysis.

<1-3. Review of performance terms> <1-3-1. Thoughts of Expressions>

Next, the expression terms suitable for the evaluation of the three-dimensionality of the face are reviewed. I took a picture of the face of the above-mentioned 21 people for the observation of the face of the person in charge of the face, and recorded the face with the characteristics of the three-dimensional and the shadow of the face. For example, "high nose", "full cheek", "face length", etc. At this time, even if there is no corresponding to the stereoscopic feeling, it is possible to freely imagine and speak without restriction. Here, FIGS. 2A and 2B show an example of a feature expression word that is recognized by a face photograph, and the number of terms actually recognized is 282.

Next, according to the meaning content, the performance terms recorded are organized according to each element. The elements use "contour", "face line", "forehead", "eye", "nose", "cheek", "mouth", "chin", "overall", "space", "eyes and noses" 13 elements of "configuration", "face", "first class". Review the content of each term, and concentrate those with similar meanings as one, and are not suitable for general performers, except those that do not correspond to the appearance of facial features including facial stereotypes and shadows, and will outline the face and cheeks. A total of 49 adjectives, such as sensitivity, were selected as tentative evaluation terms.

Also, since the point is not clear on the proper expression of the three-dimensionality of the face at this time Indeed, it also includes nine expressions for evaluating the type of face type (type feature) and the shape of the face contour. Here, Fig. 3 shows an example of an evaluation table for evaluating the three-dimensionality of the face.

Further, in the evaluation of the evaluation table of FIG. 3, for example, the evaluation table of FIG. 3 may be displayed on the display of the face classification device 10, and the evaluator or the like may input predetermined evaluation information for the face of the subject by the input device 11 or the like. And use this to evaluate. Further, the input method of the evaluation information can be set by a pointing device such as a mouse, and by selecting a cursor to a predetermined position of the predetermined evaluation item. Further, the method of inputting the data is not limited thereto, and the evaluation information may be input using, for example, a keyboard.

<1-3-2. Selection of evaluation terms>

Next, in order to review the appropriateness of the evaluation term, the face evaluation was actually performed using nine expressions of 49 adjective phrases and contour shapes. In addition, the evaluation is based on the above-mentioned 21 face image samples, and the 9 points of the hypothetical indicators of the above <1-2-2. Evaluation of the evaluation index are significant. First, use the evaluation form in Figure 3 and conduct object evaluation through four research leaders. In addition, the evaluation step has the central stage (3) as the five stages of "not biasing toward either one".

Next, the evaluation results are counted, and the average evaluation is calculated based on the object. Fig. 4 is a view showing an example of the average evaluation result of the object. In Fig. 4, the evaluation results of five (Kr, Ct, Km, Ng, Nt) among the subjects are displayed. All subjects were reviewed as shown in Figure 4, and it was judged that the project with the gap in the evaluation was an important evaluation item.

In addition, the items with large scores are "square/round", "sharp/not sharp" in the outline and face lines, and "eyes of the eyelids/eyebrows to the nose column" in the eyes. The three-dimensional sense (straight line/curve) is "long/short", "round/straight", "nose nose/not stiff" in the nose, and "skin/bone" and "flat/high" on the cheek. "The center of gravity is up/down" and "the bulge of the cheeks" is "round/square" on the chin, and "skin/bone" and "smooth/uneven" in the face as a whole, in terms of facial space. The evaluation is "forehead width/narrow" and "chin length/short".

On the other hand, the project with a small gap (the item with little change in value) is "smooth" on the face line, "width" on the forehead, "brow bone" on the eye, and "highlighted, wide" on the chin. In the whole, it is "stereo/planar".

Therefore, the items are selected based on the results, and the nine expressions of the 31 adjective phrases and the facial contour shapes that are effective in evaluating the facial features and the three-dimensionality are determined and used in the review described later.

<1-4. Correspondence between physical objects and photos>

Here, we also review the evaluation of the actual object and the photo. The same four evaluators used the photos of the same person to evaluate the nine subjects after the physical evaluation. Here, the fifth figure shows an example of the evaluation result of the physical object and the photograph, and the fifth figure shows the evaluation about the subject Ng. Here, the evaluation result of the physical object and the photograph has a tendency to be evaluated in a direction in which the unevenness is not felt as the reflectance of the reflecting plate placed under the face when the photograph is taken. Further, it is known that when photographing is performed after the reflectance is weakened, it is possible to deviate from the actual object.

Further, it is known that the shape of the nose or the shape of the face line differs visually depending on the direction of the chin, etc., and therefore it is difficult to determine only the item on the front of the face. In addition, it is also known that these items can be used in conjunction with photographs taken from oblique or sideways to eliminate the gap with physical evaluation.

<1-5. Obtaining the stereoscopic analysis index of the face> <1-5-1. Face evaluation experiment>

Here, in order to obtain an index for analyzing the three-dimensionality of the face, a review is further conducted. Specifically, 34 faces of a general test from the age of 20 to 40 years old were actually evaluated using an evaluation form based on the selected term. Here, Fig. 6 shows an example of an evaluation table based on the stereoscopic effect of the selected term. In addition, the evaluation performer includes 4 to 6 people including beauty technicians and product development leaders in addition to the above-mentioned research leader. The number of evaluators is increased to avoid each person's evaluation. However, evaluation items have a common definition and are confirmed in advance to reduce the difference in evaluation.

Further, the above evaluation can also be performed using the face classification device 10. In other words, the evaluation table of FIG. 6 is displayed on the display of the face classification device 10, and the evaluator or the like can use the input device 11 to input predetermined evaluation information by the above-described method or the like to perform evaluation.

<1-5-2. Obtained by factor analysis index>

Then, the numerical value of the evaluation result of each evaluator is input, and the average evaluation score of each subject is calculated for each item by the table calculation software or the like. Again, this value is used for factor analysis. Here, factor analysis is a statistical method that condenses information with related variables into a few potential factors. In addition, the amount of change here corresponds to the average evaluation score of each item. Therefore, by performing factor analysis, relevant projects can be aggregated and a small number of evaluation factors can be obtained.

Further, the factor analysis uses a commercially available statistical analysis software (for example, SPSS (Social Science Statistical Software) ver. 10.0J, etc.), and the factor can be performed by installing the software on a computer device such as the face classification device 10 or the like. analysis. this In addition, the factor analysis method specifies the principal factor method, and the rotation law specifies the oblique rotation (Promax) method. As a result, four factors (stereoscopic evaluation factors) were obtained. Here, Fig. 7 shows a graph of the amount of factor load that can be obtained from the result.

As shown in Fig. 7, the factor 1 (factor 1) of the factor load is composed of items such as the height of the bridge of the nose and the stereo/planar, and therefore is a factor of the stereoscopic effect (Depth). Further, the second factor (factor 2) is composed of items such as fullness, thinness, fleshiness, and boneiness of the cheeks, and thus is a factor of the flesh sensitivity. Further, the third factor (factor 3) is composed of items such as bulging/not bulging, cheek width/narrow, and the like, and is therefore a factor of width. Furthermore, the fourth factor (factor 4) is composed of items such as chin length/short, cheek length/short, and chin large/small, and therefore is a factor of length. Moreover, the cumulative correlation rate as an indicator of the degree of explanation is 63%.

Fig. 8 shows an example in which each object is drawn based on the scores of the obtained first and second factors. When analyzing the arrangement of Fig. 8, it is understood that each object is positioned, separated, and evaluated according to the characteristics of three-dimensionality and fleshiness.

Similarly, the ninth figure shows an example in which each object is drawn based on the scores of the obtained third and fourth factors. In Fig. 9, it is known that each object is positioned, separated, and evaluated according to the characteristics of the length and the width, and it is also known that the factors obtained by the analysis are common to the above <1-2-2. The speculative indicator can reflect the focus of the evaluation.

In addition, in S1 to S34 of Fig. 8 and Fig. 9, in order to remove the slack and the line of the face from the above-mentioned 34 people who are generally tested from the age of 20 to 40, the difference between the evaluation of the cheeks and the facial lines is too large. S5, S8) for analysis, and Do not display the position of the 32 subjects.

<1-6. Production of classification map>

Based on the above results, a classification map for evaluating the three-dimensionality of the face is produced. In addition, at the time of production, the obtained index is further adjusted in consideration of the convenience and versatility of the beauty consultation staff. The factor of the first factor and the factor of the flesh sensitivity of the second factor are used as the depth and the sensitivity of the flesh and blood, and are integrated into one as a composite element. Further, the width of the third factor and the length of the fourth factor are displayed in a space of a width and a length, and are integrated into one as a spatial element.

The above two elements are combined to be an evaluation index when used as a face part class. Here, Fig. 10 shows a first embodiment of a classification map classified according to the depth of the face, the sensitivity of the flesh and the space. In Fig. 10, the first axis includes a face type that includes a balance of the face space, and the second axis displays a face type formed by the depth of the face as a whole and the sensitivity of the flesh. And the coordinate space is configured with a facial image, and the facial image is set in advance in the two-axis-classified quadrant and the center of the two-axis intersection, and has a standard and ideal stereoscopic effect (for example, an average facial image) ).

Further, the first axis (the vertical axis of FIG. 10) is used to display the length and length of the forehead, cheek, or chin of the face space based on the average face set in advance. In addition, the second axis (the horizontal axis of Fig. 10) shows the degree of sharpness and sharpness of the skin and the sensitivity of the flesh, and the fullness and softness of the other side. Further, the arrangement of the vertical and horizontal axes of the present invention is not particularly limited.

Further, Fig. 10 is a diagram showing the performance (character) for displaying the characteristics of each axis for easy face evaluation (analysis), and by the computer in the face classification device 10 The drawing production can more clearly represent the facial images (photographs) of each quadrant feature. At this time, in addition to the frontal face image, the face image in the oblique direction is also created and arranged. Thereby, the problem of lowering the evaluation accuracy by the front face only can be solved.

Here, Fig. 11 shows an example of the focus of each index of the present invention at the time of evaluation. Further, the 11th (a) and (b) drawings are used to explain the balance of the face space, and the 11th (c) and (d) are used to explain the example of the depth and the flesh sensitivity.

As shown in Fig. 11, the face space analysis is performed based on the aspect ratio (width to length ratio) of the face contour, that is, the outer face line to be described later. Specifically, the width-to-length ratio is compared with the outer face line of the average face as compared with 1:1.4. If the length is less than 1.35 as shown in Fig. 11(a), it is judged to be short (wide width), and if the length is larger than 1.45 as shown in Fig. 11(b), it is judged to be long (long length). In addition, this range measures the aspect ratio of 34 faces and other external facial lines which are generally tested from the age of 20 to 40, and sets an appropriate range based on the measurement result.

Moreover, as shown in Fig. 11(c) and (d), the depth of the face and the sensitivity of the flesh are judged based on (1) the cheeky flesh feeling near the nose (2) the cheek bone (3) the ear to the face line of the chin. . In other words, when "the cheek bone is square", "the cheeks are recessed near the nose", and "the line of the face of the ear to the chin is straight", it is judged as "thin and sharp". In addition, when "the cheek bone, the cheek near the nose is round", and the "the curve of the face of the ear to the chin is curved", it is judged as "fullness and softness".

Further, the face image of the subject used in the evaluation can be stored in the auxiliary memory device 14 in advance by the imaging device 18 or the like of the face classification device 10. Then, image analysis is performed based on the pre-stored image for evaluation.

<2. Review of the adjustment method of three-dimensional sense> <2-1. Acquisition of the target three-dimensional sense>

Next, in order to adjust the three-dimensional feeling and obtain a sense of beauty, a three-dimensional sense will be evaluated as a beauty. In the review, use (a) plain, (b) natural skin tone to adjust the skin color difference, (c) various facials such as goose egg-based makeup that adjusts the three-dimensional effect to the balance of the face as a whole.

Here, the six types of models with different facial features are subjected to the basic makeup of the above types (a) to (c), and the face photographs taken by the models are used, and 20 women are 20 years old. Conduct a questionnaire survey for the subject. Here, Fig. 12 is an example of an evaluation item showing a visual effect and an overall impression on the face.

As shown in Fig. 12, 8 items for the visual effect on the face (the visual effect of the forehead, the three-dimensional sense of the forehead, the shape of the nose column, the flesh of the cheek, the length of the cheek, the line of the face, the shape of the chin, the balance of the face) The 3 items (three-dimensionality, beauty, and comfort of the entire face) of the overall impression of the overall three-dimensionality, beauty, and comfort of the face were evaluated. In addition, it is also appropriate to answer the visual effect of the face.

As a result, it was found that the balance of the face shape can be adjusted by makeup, and the three-dimensional feeling is exhibited, and the evaluation of the sense of beauty and comfort can be improved. The evaluation is based on the goose egg type foundation makeup.

Further, Fig. 13 shows an example of the evaluation result of the visual effect of the face, and Fig. 13 shows the average of the evaluations of the six 20-year-old female models. value. As shown in Fig. 13, when the three-dimensional feeling is adjusted by the goose-shaped foundation makeup, the facial evaluation can be significantly improved compared to the plain or plain-based makeup.

Further, in the detailed analysis of the goose egg-based foundation makeup, it is understood that the makeup which is similar to the contour shape of the average face is formed on the inner side of the face through the eyebrow tail.

<2-2. Review of standard balanced face>

Here, the general beauty of the face conditions are goose-shaped facial lines and gold balance. The gold balance is such that the eye position is about 1/2 of the height of the whole head, the brow is located at the position of the hairline from the hairline to the front end of the chin, and the nose is located 2/3 of the hairline.

Here, it is known that the balance of the "average face" created by averaging the size information and color information of a face image of a majority name by a conventional image synthesis technique can be approximated to the gold balance value. Further, it is known that in the production of an average face, when 10 people are used, the impression is similar to that of a face photo maker having 10 different faces (for example, Nishiya Miwa and others, "Exploring the characteristics of the average face", The 63rd Congress of the Japanese Psychological Association published a collection of papers, issued in August 1999, etc.).

Here, Fig. 14 shows an example of "average face" created using a face photograph of 40 women. The face of Figure 14 meets the aforementioned golden balance. Also, the back is balanced by the gold balance.

In addition, the 15th image is the information that removes the depth and the sensitivity of the flesh and blood from the "average face" (removing skin color information). It can be seen from Fig. 15 that the depth and the sensitivity of the flesh and blood have a considerable influence on the facial impression. In addition, it is known that when the image analysis (monochrome tone separation processing) is performed on the average face, the face line can also be used. The inner side has a goose egg shape similar to the shape of the face line.

Here, Fig. 16 shows an example of the result of image (or stereoscopic) analysis of the average face. As shown in Fig. 16, it can be seen that the inside of the face line (outer face line) 20 on the outer side of the face line corresponding to the face also has a face line which is similar to the outer face line 20 and is reduced (internal face) Line) 21. It is also known that the three-dimensional appearance of the inside of the face is also one of the new aesthetic expression methods.

<3. Adjustment of stereoscopic effect - optimal elliptical adjustment method ->

Next, in order to create a beauty method in which the three-dimensional feeling inside the face line, that is, the inner face line is adjusted to a goose egg type, an adjustment hypothesis is established. Here, the outer face line and the inner face line shape obtained from the average face are defined as "best ellipse". The outer face line is a goose egg shape (standard outer face line) with a width to length ratio of approximately 1:1.4. The inner face line of the "average face" is similar to the standard outer face line and is reduced in a predetermined ratio, and the width to length ratio is also approximately 1:1.4. Here, Fig. 17 shows an example of the best elliptical shape.

The outer face line 20 is a shape indicated by the relationship of the "width of the face: length ratio = 1:1.4". In addition, the balance of the external face line can see the focus of the adjustment of the overall space of the face. Further, as shown in Fig. 17, when the internal facial line 21 is applied to each face, the width of the left and right cheek bones of each face is first set to 1, and then the standard internal length of 1.4 is drawn based on the width. The facial line 21, i.e., the best elliptical shape, is positioned to contact the eyebrows of each face (standard on the extension line from the side of the nose through the end of the eye). Moreover, the internal facial line balance is adjusted by the flesh sensitivity The adjustment of the three-dimensional sense and space.

Further, between the outer face line 20 and the inner face line 21, a region 22 (hatched portion in Fig. 17) which gives a natural sense of depth is provided. In addition, the region 22 can adjust the shape of the outer facial line 20 and the inner facial line 21 to exhibit an adjusted area of depth when the depth is revealed.

Further, Fig. 18 is a view for explaining an example of the optimum elliptical shape. A general super ellipse can be represented by "(X/a) ⁿ +(Y/b) ⁿ =1" when the ellipse has a long radius of a, a short radius of b, and a curvature of n in the XY plane of Fig. 18. . At this time, the optimal elliptical shape is connected to the super ellipse by Y=0, and the range of the super ellipse is Y>0 (the half of the face of Fig. 18) and Y<0 (the lower half of the face of Fig. 18) Each of the variables is "flattenance b/a = 1.4 ± 0.05", and "the number of times of curvature n is 2.29 ≦ n ≦ 2.63 in the Y > 0 region and 1.75 ≦ n in the Y < 0 region. ≦ 2.00".

Further, when the shape of the inner face line 21 is adjusted as described above, the adjustment directivity differs depending on the position on the classification map of Fig. 10 described above. One of them, that is, when the face of the object is judged to be short in space balance, the adjustment makes the visual effect of the space balance longer, and conversely, when it is judged that the space balance belongs to the elder, the adjustment makes the visual effect of the space balance shorter. .

Another type, that is, when the depth and the flesh sensitivity are full and soft, the adjustment is to emphasize the bone feeling. Conversely, when the depth and the flesh sensitivity are thin and sharp, the adjustment is made to soften the bone. The foregoing is used as the adjustment method of the optimal ellipse, and the verification of the hypothesis is performed.

<4. Hypothesis verification of stereoscopic adjustment>

For the face evaluation object of the above <1-5-1. Face evaluation experiment> Use the adjustment hypothesis and use the face powder to adjust the three-dimensional makeup. The facial powder used here has two colors of focus color (light color) which is injected into the light to adjust the sensitivity of the flesh, and two colors of the modeling color of the sharp three-dimensional face by the shadow correction of the flesh sensitivity. In addition, the focus color is used in response to the skin color brightness. In addition, the makeup system is performed by a cosmetic technician.

The cosmetic product was evaluated by 4 to 6 evaluators in kind. As a result, it was found that the visual effect of the length of the cheek or the sense of flesh was adjusted, and the visual effect of the overall three-dimensionality of the face was close to the standard.

Fig. 19 is a view showing an example of the coating method of the face powder. Here, the 19th figure shows an example of the method of applying the facial powder to the subject A who has a short space balance of the face and a deep and soft flesh sensitivity.

Further, in Fig. 19, the hatched portion indicates the portion to which the focus color is applied, and the wire portion indicates the portion to which the modeling color is applied. With this adjustment, the spatial visual effect of the longitudinal side of the face can be lengthened, the cheek meat portion can be thinned, and the depth, height and low undulations can be imparted to give a three-dimensional feeling, and the internal facial lines can be presented as a goose egg type. In this way, the outer facial lines can also be presented with a goose egg shape.

Specifically, it is adjusted by using a cosmetic such as facial powder to make the space inside the internal facial line and the flesh sensitivity appear as an average face. For example, if the space is long, it will be shortened, and if the skin is noticeable, it will be full and gentle. Moreover, the outer facial lines impart a natural sense of depth by the styling color, and at this time, the inner facial lines can be optimally elliptical.

Here, Fig. 20 shows an example of the facial expression of the subject A and the facial evaluation after applying the facial powder. In addition, Figure 20 shows the evaluation of four evaluators. Mean. As shown in Fig. 20, the lines of the sharp face are sharp, and the fullness of the cheeks becomes thin and sharp, and the depth and flesh sensitivity are close to the standard. In addition, it can be seen that the standard for changing the size of the small chin and the appearance of the shape of the nose column are evaluated, and the facial space is adjusted.

<5. Type adjustment method>

Here, it is known from the above-described verification that the adjustment directivity differs depending on the position of the face determined on the map. That is, it is understood that the adjustment of the makeup can be displayed in correspondence with the axis of the drawing of the above-described Fig. 10, and the position of each face and the direction of adjustment can be displayed thereby. Fig. 21 is a view showing a second embodiment of the classification map classified according to the depth of the face, the sensitivity of the flesh and the space.

As shown in Fig. 21, the adjustment method of the stereoscopic feeling set in the quadrant classified by the two axes by the makeup adjustment is displayed on the figure. In this way, by adjusting the stereoscopic effect corresponding to the position on the map, the method of adjusting the stereoscopic feeling by the face and the makeup can be presented with higher accuracy.

Here, Fig. 22 shows an example of a makeup adjustment method for each type of face part class. First, as shown in the 22nd (a), when the balance of the face is short and the skin feels "short and thin," the makeup that balances the facial space becomes longer and the skin feels softer. . Specifically, the skin color is softened by first diluting the focus color compared to the color, giving the cheeks a full flesh and further enhancing the visual effect of the face. The focus color avoids the height of the cheek bone and fades down into the C shape along the inner side of the cheek to express a fullness. In addition, the temples are also faded to express a sense of fullness, and more faded to make the T-word part longer. The styling color system realizes that the best elliptical shape fades the forehead and the side of the chin. In addition, the protrusion of the cheek bone is also slightly diluted.

In addition, as shown in the 22nd (b), when the balance of the face space is short and full, the short and full-bodied type is used to make the visual effect of balancing the face space longer and to emphasize the makeup of the skin. Specifically, the skin color is first faded compared to the focus color to emphasize the bone feeling, and the width of the face is sculpted to make it look longer. The styling color system realizes that the best elliptical desalination causes the side portions of the contour to be elongated, and further, the cheeks are faded from the cheek bone toward the center of the face along the cheek bone, and the underside of the brow is further faded. The focus color of the cheek bone is faded to be perpendicular to the actual angle, and the fade makes the T-shaped part longer.

In addition, as shown in the 22nd (c), when the balance of the face space is long and full, the "long, full-bodied type" is used to shorten the visual effect of balancing the face space and to emphasize the makeup of the skin. Specifically, the skin color is first faded compared to the focus color to emphasize the bone feeling, and the face length is sculpted to emphasize the width line. The styling color is aware of the forehead and chin front end of the most elliptical shape with a large oval contour. In addition, the cheeks are faded horizontally from below the cheek bone toward the center of the face, and the underside of the brow is further faded. The focus color makeup is substantially perpendicular to the vertical, and the height of the cheek bone is horizontally horizontal compared to the actual angle, and the fade makes the T-shaped portion shorter.

In addition, as shown in the 22nd (d), when the balance of the face space is long and the skin feels "long, thin type", the visual effect of balancing the face space is shortened, and the skin feeling is softened. make up. Specifically, the skin color is softened by first diluting the focus color, giving the cheeks a fuller flesh and further reducing the length of the face. The focus color avoids the height of the cheek bone and fades down into the C shape along the inner side of the cheek to express fullness and fade Make the T-word part shorter. The styling color system realizes that the best oval shape greatly diminishes the hairline and the front end of the chin, and also slightly diminishes the protrusion of the cheek bone.

Further, when the face of the subject classified by the classification map has a standard and ideal three-dimensional face (for example, an average face), the makeup set in advance corresponding thereto is performed. Specifically, in the "standard, ideal type" of Figure 22(e), the focus color will fade from the height of the cheek bone to below the eyebrow, and from the forehead through the nose column to the front end of the chin. Another focus color that uses a different focus color, overlaps the previously faded focus color, and fades to penetrate the skin. The styling color fades the contour line of the face line, the front end of the chin, the hairline, etc., or the underside of the cheek bone. In addition, the other styling color different from the previously used styling color is faded to infiltrate the surrounding faded styling color or the contour portion. Skin on the inside (side of the face).

In this manner, the optimum ellipse can be adjusted by the makeup according to the face portion by displaying the stereoscopic adjustment directions of the respective types corresponding to the positions on the classification map. Further, each type of makeup adjustment method having the image shown in FIG. 5 and/or the suggested face portion may be provided to the experiment by the face classification device 10 by printing to a display screen, paper, or the like. And so on.

<6. Acceptance of adjustment method>

Finally, the subject is asked to perform a finished product evaluation by the best elliptical adjustment method of makeup. Fig. 23 is a view showing an example of the evaluation of the subject of the cosmetic product. As shown in Fig. 23, it was found that there was a high tendency to taste when evaluating the finished product of the optimal elliptical adjustment method. The preference of the finished product is 23 out of 25, and the evaluation is in the right direction (like: 11 people, slightly like: 12), 2 people who do not favor any of them, and 0 hateful respondents.

In addition, there are 15 people who like this time, and there are about 15 people who like this time. It can be seen from this that the adjustment method shows high sensitivity and is easily accepted by the general public.

<Using the makeup order of the classification map>

Next, an example of a series of makeup sequences of the subject by the face classification device 10 and a series of makeup sequences before the makeup adjustment method is displayed based on the classification result will be described using the classification chart of the present invention.

Figure 24 is a flow chart for explaining the makeup sequence using the classification chart of the present invention. First, a face image of the subject of the subject who is applying makeup is obtained (S01). Further, in S01, the front side image of the subject and the face image taken from the angle of 45 degrees can be obtained by the photographing device 18.

Next, the stereoscopic feeling of the face of the subject is evaluated based on the acquired image data (S02). Specifically, as shown in the above-mentioned 11th figure, it is compared with the standard facial line to evaluate whether the face is still short, or is skinny and thin and sharp, or is full and gentle.

Further, based on the evaluation result of S02, the evaluation of which type of subject is to be positioned in the classification chart of the present invention in Fig. 10 or Fig. 21 is performed, and the subject is classified (S03). Further, the classification system counts the evaluation information of the face of the subject according to each item as described above, and locates the corresponding classification based on the result and the preset value.

Further, the classification result is displayed on a display or the like (S04). Further, the display can display the classification map of FIG. 10 or FIG. 21, and display the evaluation result at a predetermined position of the classification map. In addition, the display can also be displayed as described above. A method of adjusting the stereoscopic effect of each classification by makeup. Further, the classification result may be printed on paper or the like to be presented to the subject or the like.

The subject or the like performs makeup corresponding thereto according to the display adjustment method or the like (S05). Specifically, the adjustment as described above causes the internal facial lines to exhibit an optimal elliptical shape. Further, the method of performing the above-described adjustment is stored in advance according to each of the classifications as shown in Fig. 19, and an adjustment method corresponding to the classification result of the subject can also be displayed.

As described above, applying makeup using an internal facial line similar to a standard facial contour (external facial line) can be used as a guideline for the aesthetics of the subject to be aesthetically pleasing.

As described above, according to the present invention, the three-dimensionality of the face can be classified with high precision. Moreover, the subject can be appropriately adjusted according to the type of classification, and specifically, a method of classifying the three-dimensionality of the face by the length of the face and the sensitivity of the whole face (especially the cheek), and Prompt to use the classification map of the purpose and adjust the stereoscopic effect by the classification type.

Also, the adjustment uses an internal face line similar to a standard face outline (external face line). Thereby, the three-dimensional feeling of the face can be classified with high precision according to each face type, and the overall harmony of the face can be obtained, and it is related to the high evaluation of the beauty and the comfort.

Moreover, the classification map can be used as a makeup tool for makeup, and it is also possible to indicate in which direction the makeup is to be oriented as a guideline for expressing the target image, and it is necessary to know the appropriate makeup adjustment method without requiring special skill. .

Thereby, the visual effect of the outline can be adjusted to "goose egg type" at the time of basic makeup, and the whole face can be reconciled. In other words, the "outer face line" of not only the outline can be made into a goose egg type, and the "internal face line" on the inner side of the face is also a so-called "best elliptical" face type of a goose egg type.

That is, by using the present invention, the target makeup can be displayed by the internal face line (super ellipse) even if the stereoscopic analysis is not performed. The makeup target with the internal facial line as a guiding principle does not depend on the three-dimensional characteristics of various faces and can be determined by one meaning, so it is a universal method applicable to most people.

Moreover, in the makeup adjustment of the present invention, the stereoscopic feeling of the face can be classified in advance, and the adjustment of both the shadow and the stereoscopic effect can be used depending on the facial features. Therefore, the stereoscopic effect can be appropriately adjusted.

The above is a detailed description of the preferred embodiments of the present invention, and the present invention is not limited to the specific embodiment, and various changes and modifications can be made within the scope of the invention as disclosed in the appended claims.

The present invention claims priority from Japanese Patent Laid-Open Publication No. 2005-348320, filed on Dec. 1, 2005, the entire disclosure of which is hereby incorporated by reference.

10‧‧‧ Face classification device

11‧‧‧ Input device

12‧‧‧ Output device

13‧‧‧ drive

14‧‧‧Auxiliary memory device

15‧‧‧ memory

16‧‧‧CPU

17‧‧‧Network connection device

18‧‧‧Photographing device

19‧‧‧Recording media

20‧‧‧External facial lines

21‧‧‧Internal facial lines

22‧‧‧ Area (slashed part)

A‧‧‧ target

B‧‧‧System Bus

Fig. 1 is a view showing an example of the device configuration of the face classification device of the present invention.

Fig. 2A shows an example of a feature expression term (1) that is recognized by a facial image.

Fig. 2B shows an example (2) of the feature expressions that are recognized by the face image.

Fig. 3 is a view showing an example of an evaluation table for evaluation of the three-dimensionality of the face.

Fig. 4 is a view showing an example of the average evaluation result of the object.

Fig. 5 is a view showing an example of evaluation results of physical objects and photographs.

Fig. 6 is a view showing an example of a stereoscopic evaluation table based on the selected language.

Figure 7 shows the factor load.

Fig. 8 shows an example in which each object is drawn by the number of scores of the first and second factors obtained.

Fig. 9 shows an example in which each object is drawn by the number of scores of the third and fourth factors obtained.

Fig. 10 is a view showing a first embodiment of a classification map classified according to the depth of the face, the sensitivity of the flesh and the space.

The eleventh (a) to (d) diagrams show an example of the focus of each index of the present invention at the time of evaluation.

Fig. 12 is an example of an evaluation item showing a visual effect and an overall impression on the face.

Fig. 13 shows an example of the evaluation result regarding the visual effect of the face.

Fig. 14 is an example of an "average face" made using a face photograph of 40 women.

Figure 15 shows the information on the depth and the sensitivity of the flesh and blood from the "average face".

Fig. 16 is a view showing an example of the result of the shadow analysis of the average face.

Figure 17 shows an example of the best elliptical shape.

Fig. 18 is a view for explaining an example of the optimum elliptical shape.

Fig. 19 is an example of a record showing the application method of facial powder.

Fig. 20 is a view showing an example of the face of the subject A and the face after applying the face powder.

Fig. 21 is a view showing a second embodiment of a classification map classified according to the depth of the face, the sensitivity of the flesh and the space.

The 22nd (a) to (e) figure shows an example of the makeup adjustment method for each type of face part.

The 23rd (a) to (b) figure shows an example of the evaluation of the subject of the cosmetic product.

Figure 24 is a flow chart showing the makeup sequence using the classification chart of the present invention.

Claims (15)

A face classification method for classifying a stereoscopic feeling of a face of an experimenter, characterized in that it comprises: an evaluation step, which is based on at least a length or a shortness of the face space and a depth of the whole face, and a flesh sensitivity Presenting the face of the above-mentioned subject with two indicators of thinness or sharpness and fullness and softness; and the classification step, which is based on the results of the evaluation steps described above, and is positioned on the face space that has been classified in advance and is long The overall depth of the face, the sensitivity of the flesh and blood are thin and sharp, the face space is long and the depth of the face is full, the flesh sensitivity is full and soft, the face space is short and the overall depth of the face is thin and the flesh sensitivity is thin. The sharp type, the face space is short, and the depth of the whole face and the flesh sensitivity are at least four types of fullness and softness. The depth of the whole face and the sensitivity of the flesh and blood are based on the cheek feeling near the nose. The cheek bones and ears to the chin's face line to judge. The face classification method according to the first aspect of the patent application, wherein the evaluation step uses the average face set in advance and each of the four types of the aforementioned types to evaluate the face of the subject. The face classification method according to claim 2, wherein the foregoing evaluation step reflects a ratio of a face width of a length and a width balance of the forehead, the cheek or the chin of the face space according to the average face evaluation, and The aforementioned depth and flesh sensitivity were evaluated by the degree of sharpness or the degree of softness. The face classification method according to any one of claims 1 to 3, further comprising a display step of displaying the face type classified by the classification step in the respective quadrants of the plane coordinate space and having the foregoing average In the face and the above-mentioned four types of drawings, in order to make the three-dimensional feeling of each of the above four types closer to the three-dimensional feeling of the average face, the method of adjusting the stereoscopic feeling of each of the above four types of makeup is displayed. . The face classification method of claim 4, wherein the adjustment method is similar to the shape of the contour of the average face, and the width is reduced by a predetermined ratio to form a shape having a length ratio of 1:1.35 to 1.45. . A face classification device for classifying a stereoscopic feeling of a face of an experimenter, comprising: a photographing mechanism for photographing a face image of the subject; and a control mechanism capable of being photographed Facial images, at least according to whether the face space is long or short and the depth of the whole face, the flesh sensitivity is thin and sharp, or the fullness and softness of the two indicators, positioning in the face space that has been classified in advance and the face is overall The depth and flesh sensitivity are thin and sharp, the face space is long and the depth of the face is full, the flesh sensitivity is full and soft, the face space is short and the whole face is deep, and the flesh sensitivity is thin and sharp. a controller having a short facial surface and a depth of the whole face, and a flesh sensitivity representing at least four types of fullness and softness; and a display mechanism for displaying the classification result classified by the control mechanism Four quadrants have been located in each quadrant of the plane coordinate space In the type of figure, the depth of the entire face and the sensitivity of the flesh and blood are judged based on the cheeky flesh feeling near the nose, the cheek bone and the face of the ear to the chin. The face classification device according to claim 6, wherein the control means uses the average face set in advance and each of the four types of the above-mentioned types to evaluate the face of the subject. The face classification device of claim 7, wherein the control mechanism evaluates a ratio of a face width of the length and width balance of the forehead, cheek or chin as the face space according to the average face. The depth and the flesh sensitivity are evaluated by the degree of sharpness or the degree of softness. The face classification device according to claim 7, wherein the display means adjusts each of the four types of makeup according to the face shape according to the classification, in order to be closer to the visual effect of the three-dimensional feeling of the average face. The method of adjusting the stereoscopic effect of the type. The face classification device according to claim 9, wherein the adjustment method is adjusted according to a shape similar to the contour of the average face, and the width reduced by a predetermined ratio is formed to have a length ratio of 1:1.35 to 1.45. A face classification map includes: a first axis, which displays a face type that is long or short according to a face space; and a second axis that displays a depth according to the entire face and a sensitivity of the flesh. The face is thin and sharp, and the face type is full and soft; and the coordinate space is configured with a face image, and the face image is set in advance in the quadrant of the two axes and the center of the two-axis intersection, and has The standard and ideal three-dimensionality, the depth of the whole face and the sensitivity of the flesh and blood are judged according to the cheeky flesh feeling near the nose, the cheek bone and the face of the ear to the chin. For example, in the face classification chart of claim 11, wherein the first axis system indicates that the face of the face, the cheek or the chin is balanced with respect to the face of the face space. The ratio of the width is large; while the second axis shows the depth and the sharpness of the flesh sensitivity, and the other side shows the fullness and softness. For example, in the face classification chart of claim 11 or 12, the method of adjusting the stereoscopic effect by the makeup is displayed on the quadrant classified by the two axes to approximate the stereoscopic feeling of the average face. A face classification program that enables a computer operation to perform the following steps: the evaluation step is based on at least two aspects of whether the face space is long or short and the depth of the face as a whole, and the flesh sensitivity is thin or sharp or full and soft. The face of the subject is evaluated; and the classification step is based on the result of the evaluation step, and the face space that has been classified in advance is long and the depth of the whole face is thin, and the sensitivity of the flesh is thin and sharp. The face space is long and the depth of the whole face is deep, the flesh sensitivity is full and soft, the face space is short and the overall depth of the face, the flesh sensitivity is thin and sharp type, face The space of the part is short and the depth of the whole face and the sensitivity of the flesh and blood are at least four types of fullness and softness. The depth of the whole face and the sensitivity of the flesh and blood are based on the cheeks and cheekbones near the nose. Judging from the ear to the face of the chin. A recording medium, which records a face classification program and can be read by a computer, can cause the computer operation to perform the following steps: the evaluation step is based on at least the length or the length of the face space and the depth of the whole face and the sensitivity of the flesh. Thin and sharp is still two indicators of fullness and softness, and the face of the above-mentioned subject is evaluated; and the classification step is based on the result of the evaluation step, and the face space that has been classified in advance is long and the face is presented. The overall depth and flesh sensitivity are thin and sharp type, the face space is long and the depth of the whole face is deep, the flesh sensitivity is full and soft, the face space is short and the whole face is deep, and the flesh sensitivity is thin and sharp. The face space is short and the depth of the whole face and the flesh sensitivity are at least four types of fullness and softness. The depth of the whole face and the sensitivity of the flesh and blood are based on the cheeks and cheeks near the nose. The bones and ears are judged by the lines of the face of the chin.