TW201028963A - Evaluation method of skin color, evaluation apparatus of skin color, evaluation program of skin color and recording media thereof - Google Patents

Abstract

An evaluation method of skin color is disclosed, which evaluates the skin color according to the inputted face region image, and comprises a dividing step, an evaluation step of skin color distribution, and an image frame constructing step. The dividing step divides the face region image into predetermined regions according to a first characteristic point constructed by at least 25 predetermined locations in the full face region of the face region image and a second characteristic point setup by the first characteristic point. The evaluation step of skin color distribution constructs the mean value of the skin color distribution by using at least one of, L * a * b *, Cab * of L * a * b * color table, three exciting value of X, Y and Z of XYZ color table, each RGB value, color phase H, lightness V, chromaticity C, melanin number and hemachrome number. The method evaluates the skin color according to the measurement results. The image frame constructing step displays the aforementioned measurement results or evaluation results on a screen.

Description

201028963 VI. INSTRUCTIONS OF THE INVENTION: TECHNICAL FIELD OF THE INVENTION [0001] The present invention relates to a skin color evaluation method, a skin color evaluation device 5, a skin color 3 evaluation program, and a recording medium on which the program is recorded. In particular, there are a skin color evaluation method capable of evaluating skin color with high accuracy, a skin color evaluation device, a skin color evaluation program, and a recording medium on which the program is recorded. Lu [First * Hu's old U Xuan] Background of the invention 10 _2] The conventional method of 'diagnosing and evaluating facial skin color is to use the color leaf to measure the color of the face 1 to several points, and the result is used as the skin color data. Method of evaluation (refer to Non-Patent Documents 1 and 2). The advantage of this method is that it can be calculated by comparing the distribution or average of Japanese women in this part by taking a lot of information on the 'stationary part.' [0003] Here, Fig. 1 shows an example of a skin color evaluation by a conventional method. Further, in the example of Fig. 1, the vertical axis shows the brightness and the horizontal axis shows the hue. In the example shown in Fig. 1, from the majority of the measurement results, the average of the skin color under the cheeks of Japanese women is 11 and 95% of the skin color under the cheeks of the woman's cheeks is extended to 12, one is measured. The tester measures the skin color under the cheeks, and judges which position of the far measurement is relative to the average value 11 or the region (for example, 95% trusts the outer side or the inner side of the ellipse 12), thereby performing personal skin color evaluation. In addition, the skin color comparison before and after the use of the same person's cosmetics and the evaluation of the skin color comparison with others are also carried out. [0004] However, there is a problem that "the data of only a few points in the face cannot represent the skin color of the face 3 201028963" and the need to obtain various parts of the face at the same time to evaluate. Therefore, a method of performing skin color evaluation using image data photographed by a digital camera has been developed recently (for example, refer to Non-Patent Documents 3 and 4). The advantage of this method is that the entire face is divided into hundreds of places, and the skin color of these parts can be grasped. Non-Patent Document 1: Chai Gu and others, "Study on skin color and makeup effect (1st report) - Development of skin color measurement method and study of makeup effect _", makeup technique, ν〇 1.17, No. 2. 1983. Non-Patent Document 2 · Shibatani et al., "Measures of Ag ing Effect of Facial Color Distribution and Applications j , J. Soc. Cosmet. Chem. Japan. V9 > nl ' 1985. Non-Patent Document 3 : Was And others, "A new attempt to reduce the color of skin and uneven skin color in women's face - Development of polarized image analysis system and application of skin color unevenness evaluation", Total makeup, Vol.26, No.2, 1992. Non-patent Document 4: Caisey and others, "The female skin color and makeup strategy of different people's groups", FRAGRANCE, JOURNAL 2007-4. C SUMMARY OF THE INVENTION 1 The invention reveals the subject to be solved by the invention [0005] However, 'as is known As shown in the method, when using images shot with a digital camera, even if they are the same (X, y) coordinates, the relative positions in the face are different, even if they are the same person, there is a position every time. Different situations. Therefore, even with the comparison or difference of the data taken by each punctuation point, it is impossible to perform high-accuracy evaluation. 201028963 [_6] In addition, it is not possible to calculate the skin color value as a beneficial indicator, and the result is stagnant in the calculation of the data of the average or average part of the face. Furthermore, as described above, the difference between the face and the color of the face is 2 = ^, and the comparison of the skin color of the skin before and after the use of the homogenized product is compared with the skin color of the other person. The images produced errors and were not able to be evaluated with high accuracy. [0007] The present invention is invented in view of the above-mentioned problems, and 10 15 20 is used for evaluating the skin color evaluation method of high skin color with high accuracy, waste system placement, and skin color evaluation. _Color evaluation is used to solve the problem [_8] In order to solve the above problems, this _ (4) has a means to solve the problem. [0009] The present invention is a skin color evaluation method for evaluating a skin color from a face region of a wheeled person, which has a segmentation step, a wide color eight (four) estimation step, and a screen generation (four): the wording is repeated : Image: Γ: = at least 25 positions set first ^ 吏 Use the second feature point set by the first feature point to be divided into predetermined areas

2; The skin color distribution evaluation step is in (4) the segmentation step age segmentation field uses the LW color system u*, a*, b*, Cab*, xY^M two stimulation values X, Y, Z, RGB values At least one of hue H, lightness V, chroma C, melanin amount and amount of prime! The 'shadow distribution of the average value generated' is evaluated based on the measurement result; the screen generation (4) displays the aforementioned measurement result or evaluation result on the screen. _〇] Thereby, the three stimulus values X, Υ, Ζ of the L*, a*, b*, Cab*, and XYZ color systems of the L2010 color system are used for each region divided by the above-described dividing step. , RGB values, hue Η, brightness ν, chroma c, melanin amount and hemoglobin amount at least tamping, using the average skin color distribution for skin color evaluation 'so that the skin color can be evaluated with high precision. 5 [〇〇11] Further, the aforementioned dividing step sets a plurality of face lines of at least 25 of the forehead of the face/image, the vicinity of the left and right eyes, the nose, the mouth, and the underline. [0012] Thereby, since the skin color can be different depending on the face portion, the feature point can be set, and the evaluation of the region having the different skin color can be performed with high accuracy. [Thank 3] The above-mentioned _step_ is divided into three regions in which the third feature point of the second feature point and the second feature point are surrounded by the feature points of ±. [_4] By selecting the most appropriate area by 'selecting feature points', the correct evaluation of the skin color distribution can be performed. 15 _5] In addition, the skin color distribution evaluation step is a three-stimulus value χ, γ, z, Γ value, hue H, lightness V, chroma C, of the LW, Cab*, χγζ color system using the color system. The amount of melanin and the amount of hemoglobin to v 1 'the skin color divided according to the average of the divided regions, the face is processed to form a plurality of (4) images of the synthetic silk, and the average face is evaluated from the response. The skin color distribution of each region of the segmentation is based on the face Lj6] by which the plurality of face shapes can be synthesized by riding, and the average face is averaged and evaluated by the wealth. [0〇ΐ7] In addition, the skin color distribution evaluation step silk m* color 20 201028963 is the LW, Cab*, XYZ color system three stimulation values χ, γ, z, rgb values, hue H, brightness V, The average value of at least i of the chroma C, the amount of melanin, and the amount of hemoglobin is evaluated based on the average value obtained in the similar regions. 5 [〇〇18] Thereby, it is easy to perform skin color distribution evaluation based on pre-set characteristics and pooling in similar areas. [0019] Further, the aforementioned skin color distribution evaluation step generates a pre-set skin color distribution, a past skin color distribution, a skin color distribution of another person, an average value of a complex skin color distribution, and a skin color distribution of the aforementioned collection area. The evaluation of the skin color distribution is performed by comparing the skin color distribution to the target. [0020] By this, it is possible to obtain the average data of each person belonging to a certain category (age, occupation, gender), the information of an ideal candidate such as an entertainer, the personal past data, the data of others, and the like. The difference between the two, so 15 help advice when selling cosmetics. Further, the present invention is a skin color evaluation device for evaluating a skin color from an image including a face region of a person to be input, which has a segmentation mechanism, a skin color distribution evaluation, and a screen generation mechanism, which is based on the aforementioned The i-th feature point formed by at least 25 positions in advance of the entire face region of the image and the second feature point set using the first feature point are divided into predetermined regions, and the skin color distribution evaluation mechanism is configured by the above-described dividing mechanism Each region of the segmentation uses L*, a*, b*, w, χ of the LVb* color system, the stimulus value X, Υ ' Ζ, RGB values, color sensation, brightness ν, chroma ^ melanin amount and At least (10) of the amount of erythropoietin, the skin color score of the average value is generated. 7 201028963 The cloth is 'evaluated according to the measurement result; the screen generating mechanism displays the aforementioned measurement result or evaluation result on the screen. [0022] Thereby, the L W color system u W is used in each of the regions divided by the division mechanism. *, 三 ζ ζ ζ 之 三 三 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 ! ! ! The skin color distribution is evaluated using the skin color distribution of the average value, so that the skin color can be evaluated with high precision. [0023] Further, the rhyming means sets at least 25 of the face lines of the forehead of the entire face image, the vicinity of the left and right eyes, the nose, the mouth, and the face line under the eyes. [0024] Thereby, since the skin color can be known from the face portion, it is considered that the feature points are set, and the evaluation of the region having the different skin color can be performed with high accuracy. _5] The above-mentioned separation structure is divided into three parts: the _^ general point package _93 areas. By using the 'Wei Wei point' and setting the optimum area, you can make a correct assessment of the wide color distribution. _7] and 'the aforementioned skin color distribution evaluation mechanism mine makes the private *a*b* color system LW, ^, χ ζ ζ ζ 三 三 X X X X X : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : Degree c, amount of melanin and amount of hemoglobin, = skin color divided according to the average of each region of the fractal, face (4) variable processing, multiple face image prepared by the silk, and the average face segmentation from the average The distribution of skin color in each region corresponds to the assessment. 20 201028963 [0028] Thereby, the shape of the composite complex ore portion can be gradually processed, and the average face averaged by the face shape can be evaluated with high accuracy. [0029] Further, the skin color distribution evaluation means obtains three values of the three stimulus values χ, γ, z, and rgb 5 of the LW, Cab*, and XYZ color systems of the L*a*b* color system, and the hue H, The average value of at least one of the diurnality V, the chroma C, the melanin amount, and the remaining amount is evaluated based on the average value obtained in the similar regions. • _G] By this, it is easy to perform skin color distribution evaluation based on pre-determined features and collection in similar areas. 10 [GG31] In addition, the skin color distribution evaluation mechanism generates an ideal skin color distribution, a past skin color distribution, a skin color distribution of another person, an average value of a complex skin color distribution, and a skin color distribution of the aforementioned collection area to « One less is used as the comparison skin color distribution, and the comparison with the personal data of the subject is performed to evaluate the skin color distribution. 15 [GG32] By means of the average data of each person who can obtain a certain category (age, occupation, gender), the artist (4) the information of the person who wants to be selected, the information of the individual, the others The difference between the data and the like, so it helps to sell cosmetics. [0033] In addition, the present invention (4) estimates the skin color evaluation program from the image of the face region of the input person, which causes the computer to perform the segmentation step, the skin color distribution evaluation step, and the face generation step, the segmentation step The skin color distribution evaluation step is based on a third point composed of at least 25 positions set in advance on the entire face region of the image, and a second feature point set using the first feature point. In the 201028963 area divided by the above-mentioned dividing step, the L*,**, b*, Cab*, and XYZ color systems of the L*a*b* color system are used for the three stimulus values X, Υ, Ζ, and RGB. At least one of the value, the hue Η, the brightness V, the chroma C, the melanin amount, and the hemoglobin amount, and the skin color distribution that generates the average value is evaluated according to the measurement result; the screen generation step is 5 the aforementioned measurement result or evaluation The result is shown on the screen. [0034] By the use of the LW color system L*a*b*, Cab*, χγζ color system, the three stimulus values X, Y, Z, RGB values are used in each region divided by the above-mentioned dividing mechanism. At least one of the hue, the brightness V, the chroma c, the melanin amount 10, and the hemoglobin amount, and the skin color distribution of the average value is used to evaluate the skin color, so that the skin color can be evaluated with high precision. Further, by installing the program, the skin color distribution evaluation of the present invention can be easily realized by a general-purpose personal computer or the like. [0035] Further, the present invention is a computer-readable recording medium on which a skin color evaluation program is recorded, which is a method for evaluating a skin color from an image including a face area of a wheel that is subjected to a segmentation step. , skin color distribution, 5 leveling step, and facet generating step'. The dividing step is based on the second feature set by using at least 25 (four) of the face area of the image to be used first. The feature points are divided into predetermined regions; 1 The skin color distribution evaluation step is performed in each of the above-mentioned (4) steps (4), using the L*a*b*, 20 佶γ v, χγΖ, and the three-stimulus Z, RGB of the color system Each value, hue Η, brightness V, chroma C, quantity, and quantity 肤色 1 生成 生成 生成 生成 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色 肤色Picture. _6] This can be used to record the media to install the skin color evaluation program on its 10 201028963. Further, by installing the program, the skin color distribution evaluation of the present invention can be easily realized by a general-purpose personal computer or the like. Effect of the Invention [0037] According to the present invention, skin color can be evaluated with high accuracy. 5 BRIEF DESCRIPTION OF THE DRAWINGS [0038] Fig. 1 is a diagram showing an example of skin color evaluation by a conventional method. Fig. 2 is a view showing an example of the functional configuration of the skin color estimating device of the present embodiment. 10 Fig. 3 is a view showing an example of a hardware structure in which the skin color evaluation processing of the present embodiment can be realized. Fig. 4 is a flow chart showing an example of the skin color evaluation processing program of the embodiment. Fig. 5 is a view showing one of the feature points and the divided regions of the present embodiment. Fig. 6A shows an example (1) of the positional relationship in the face corresponding to the 109 feature points in Fig. 5 described above. Fig. 6B is a diagram showing an example (2) of the positional relationship in the face corresponding to the 109 feature points in Fig. 5 described above. 20 Fig. 6C shows an example (3) of the positional relationship in the face corresponding to the 109 feature points in Fig. 5 described above. Fig. 6D shows an example (4) of the positional relationship in the face corresponding to the 109 feature points in Fig. 5 described above. Fig. 7A shows an example (1) of a combination of the features 11 201028963 constituting the respective regions corresponding to the above-mentioned Fig. 5 . Fig. 7B shows an example (2) of a combination of feature points constituting each region corresponding to the above fifth graph. Fig. 7C shows an example (3) of a combination of five points constituting the features corresponding to the respective regions of the fifth drawing. Fig. 8 shows an example of another feature point setting.

Fig. 9A is a diagram (1) for explaining a flow of face division of an image obtained by photographing. Q Fig. 9B is a diagram (1) for explaining the flow of the face segmentation of the image obtained by the shooting. Fig. 9C is a diagram for explaining an example of the flow of face division of an image obtained by photographing (3). Fig. 10A shows an example of the skin color distribution for comparison (1). ^ Figure 10B shows an example of the skin color distribution for comparison (2). 15 Figure 10C shows an example of a skin color distribution for comparison (3).

The 10th figure shows an example of the skin color distribution for comparison (4). Q Figure 10E shows an example of the skin color distribution for comparison (5). Fig. 11 is a diagram for explaining a comparative example of the distribution of the skin color distribution of the difference. Figure 12 shows an example of an area classified into groups. Fig. 13 is a view showing an example of the skin color characteristics of the main components of the respective groups in Fig. 12 and the area numbers constituting each group. Fig. 14 is a diagram showing an example of the evaluation result of the skin color distribution. Fig. 15 shows an example of the hue and brightness V for explaining the difference between the groups. 12 201028963 Figure 16A shows an example of a straight square corresponding to the hue of Figure 15. Fig. 16B shows an example of a histogram corresponding to the brightness v of Fig. 15. 5

10 15

C mode j is the best mode for carrying out the invention [0040] <Summary of the present invention> The present invention can divide the input face image by a predetermined method, and the sputum face color* cloth, and the average face standard The face is replaced with a wide color, and the light can be entered into the light only to exclude the appearance of the skin shape of the face. Moreover, the average value of each of the persons belonging to the _ or the difference between the two may be taken. [0041] Hereinafter, using the drawing, the skin color evaluation device, the skin color evaluation program, and the recording of the Cheng Wuyue recording media suitable for implementing the skin color evaluation method of the present invention will be described. [0042] Cut Moon & '«ΤΙ* 衣直一用图式' The functional scale 20 of the skin color evaluation device of the present embodiment is shown as a phase 12 diagram showing one of the components of the skin color evaluation device of the present embodiment. The skin color evaluation device 2G shown in Fig. 2 is configured to be different from each other, the abuse mechanism 21, the output mechanism 22, the storage mechanism 23, and the face division mechanism [0043: evaluation mechanism 25, screen generation mechanism %, and control mechanism 27. The image of the domain uHH21 is input by (4) the face segmentation instruction including the face of the user's face, the skin color distribution evaluation instruction, and the start or end of various indications of the screen creation instruction of 201028963. Further, the input unit 21 also has a function of inputting a captured portion image including a user W (four) performing an operation by a photographing means such as a digital camera. The output unit 22 performs display and output of the content input by the input unit 2 or the content executed by the input unit. Further, the output mechanism 22 is constituted by a display, a speaker, or the like. Further, the output unit 22 may have a function of a printer or the like. In this case, the input image, the segmentation result of the face region, the skin color distribution evaluation result, and the like may be printed on the paper by the screen generating means 26 or the like. Printing media, but for users and so on. Further, the storage unit 23 stores various kinds of materials such as the face division result of the face division mechanism 24, the skin color distribution evaluation by the skin color distribution evaluation unit 25, and various screen generation results of the facet generation mechanism 26. In addition, the depository unit 23 may add the name of the slot, the date and time, the name of the person (subject) or the age of the person (15th generation), gender, when storing various images or various processing results. Personal identification information such as ethnicity is stored. Further, the storage unit 23 can read various stored materials as needed. Further, the face division unit 24 divides the predetermined area by a method of setting the input image in advance. Specifically, the face segmentation mechanism 24 faces the face. P is preset in advance; at least 3 points are selected from the set feature points, and the division method of the region surrounded by the selected feature points is used, and the appropriate region is divided for evaluation of the skin color distribution. The details of the face division method of the present embodiment I will be described later. The skin color distribution evaluation unit 25 uses the LW color system L* 14 201028963 for each of the divided regions in accordance with the segmentation result obtained by the face segmentation unit 24.

Cab XYZ color system tristimulus values X, Υ, Z, RGB values, hue clarity \^ chroma c ' at least one of melanin amount and hemoglobin amount' measured skin color distribution measurement. In addition, the skin color distribution mechanism fiber measurement results, and the skin color distribution is evaluated. 5 [Flip] In addition, the above-mentioned amount of melanin and hemoglobin, such as phase and others, the development of new speckle measurement method, makeup technique, ν〇ι.35, version*, and spirit, can be used in advance. The method of measuring the relationship between the amount of melanin and the amount of hemoglobin and the tristimulus values X, Y, and Z is obtained, but is not limited to this method. [0049] In addition, the skin color distribution evaluation mechanism % according to the measurement results of the skin color distribution of 10 cloths by computer graphics technology - the gradual processing, the synthesis of a plurality of face images in advance 'and the average face from the face shape averaged The skin color distribution of each of the divided regions is correspondingly evaluated. [0〇5〇] Further, the skin color distribution evaluation unit 25 generates a predetermined ideal skin color 15 cloth, a past skin color distribution, a skin color distribution of another person, and a complex skin color distribution using a plurality of image data and skin color distribution data stored in advance. At least one of the average value and the skin color distribution towel in the region of each predetermined area is used as a comparison skin color distribution, and the generated skin color distribution is used to obtain a difference from the acquired data of the evaluation target image, thereby using L* a*b* table ^ The three stimulus values of the l ab*, cab*, XYz color system x, y, Z rgb 2 〇 each value, hue 明, lightness V, chroma c, melanin amount and gray pigment At least one of the quantities is collected (grouped) in each of the similar regions of the average value, and evaluated based on the skin color distribution (skin distribution map) of the collected regions. [_1] In addition, the skin color distribution is compared, and the personal face recognition information added to the stored data is used to generate the average face of the age group 15 201028963 skin color distribution or human skin color distribution. Further, the skin color distribution evaluating unit 25 may compare the skin color distribution with the skin color distribution prepared in advance and perform evaluation. [0053] The screen generating means 26 generates an instruction to the user (in accordance with the input instruction of the input means 2) by the user or the like, the image data obtained by the dividing means 24, or the evaluation result obtained by the skin color distribution evaluating means 25, etc. The prompt screen is output by the output mechanism 22. In addition, the facet generating mechanism 26

It is also possible to color or generate a chart or a table from a predetermined area from the acquired image or numerical value. 1〇[0054] The control unit 27 performs control of the entire components of the evaluation device 20. Specifically, the control unit 27 performs face segmentation processing, skin color distribution evaluation processing, and screen according to an instruction of the user to the input mechanism 21 or the like. Generate various controls such as processing. 15 <skin color evaluation device 20: hardware structure>

Here, for each configuration of the skin color evaluation device 20, an execution program (skin color evaluation program) that can execute various functions of the computer is generated, and the execution program is installed on a general-purpose personal computer, a server, or the like, and the skin color of the present invention can be realized. Evaluation processing, etc. [0056] Here, a hardware configuration example of the skin color evaluation processing of the present embodiment can be realized by using a drawing. Fig. 3 is a view showing an example of a hardware structure in which the present embodiment can be realized. [0057] The computer body of FIG. 3 is configured to have an input device 31, an output device 32, a drive device 33, an auxiliary memory device 34, a memory device 35, and a memory device for performing various controls (Ceca i Pr. Coffee Uni (10) The network connection device 37 is connected to each other by the system bus bar B. [0058] The input device 31 has a keyboard for a user or the like, a pointing device such as a mouse, and the like for inputting a program from a user or the like. The wheeling device 31 has an input unit for inputting a part or all of the image of the face of the employed person photographed from a photographing mechanism such as a camera. [0059] The output device 32 has a display operation for display. The present invention provides a display for various windows or materials required by the computer body, and can be carried out by the CPU 36 (4), the display of the thinning (10) or the result, etc. 1〇[面] In addition, the 'input device 31 and The wheel-out device 32 may be an input/output mechanism such as a touch panel. In this case, the user's finger or a pen-type input device is used to touch the predetermined position and input. _1] This hair The executable program installed on the computer body is provided by a portable recording I5 media 38 such as USB (Universal Serial Bus) memory or cd r〇m. § The recorded recording medium 38 can be set in the driver package 433. The execution program included in the record (4) 38 is mounted on the auxiliary memory device 34 from the recording medium by the drive device 33. [0062] The auxiliary memory device 34 is a storage device such as a hard disk, and stores the execution program of the present invention or is controlled by a computer. The program can perform 20 input and output as needed. [0063] The memory device 35 stores an execution program and the like which are retrieved from the auxiliary memory device by the CPU 36. In addition, the memory device 35*R〇M (Read 〇nly)

Memory) or RAM (Random Access Memory). [0064] The CPU 36 controls the various programs, the input of the data of each hardware component, or the computer as a whole, according to an execution program such as an OS (〇perating System) and a program stored in the memory device 35. The processing can realize the processing of the color distribution evaluation. Further, various kinds of assets and the like required for execution of the program can be obtained from the auxiliary storage device 34, and the execution result and the like can be stored. [0065] The network connection device 37 is connected to the communication network, etc., from the secret money connected to the communication network, or the execution of the program. ° In addition, by using the rest mode, you can "pass the body of the invention. Estimate the treatment. Realize this [0067] <Glossy evaluation processing program> The picture shows the skin color evaluation processing program of this embodiment. Fig. 4 is a flow chart showing an example of the wide color evaluation processing program of the present embodiment. [〇_4 The color skin evaluation processing of the fourth image is first input to the evaluation object including the face photographed by the photographing (4). Image_), the image of the input face is divided into a predetermined number by the predetermined setting method, and the image obtained by S01 can be used.) This is a camera with a full camera and the like. The image, the field and others, for example, can use the spotting device that uses the image development, makeup technique, V〇1.28, N〇2 10 spotting quantification system using “舛 and others, etc.” The image taken, ', 94·, etc., revealed in the box, ==== =_ Machine lighting feet. Oh, but in front of the lighting box

18 201028963 A number of odont bulbs, so that the camera captures the face and obtains the captured facial image. Further, the image used in the present invention is not particularly limited, and an image taken in a general illumination environment such as a fluorescent lamp can also be used. ❿ 10 15 20 _Throat, generate skin color from images divided into predetermined areas: coffee) Use a variety of pre-stored materials to generate a comparison skin color distribution (10) sentence. Further, the evaluation of the skin color distribution profile (S06) is carried out by using the skin color distribution obtained by the processing of SG3 and the skin color distribution obtained by the comparison between the skin color distribution and the skin color distribution. _1] In addition, the evaluation result obtained by the unique processing is generated for the use = display screen (_, output generated by the face (evaluation, etc. XS08). [〇〇% is here] to determine whether to continue the skin color evaluation (s〇9 When the skin level estimation is continued (YES in the chat), the process returns to S02, and the division by the different division method is performed, and the processing described later is performed. When the treatment of S09 does not continue the skin color evaluation ( In the step of 9, for example, N〇), the processing is terminated.] By the shape of the interrogation unit, the skin/body can be evaluated with high precision. The segmentation method of the present invention divides the input facial image into predetermined numbers. In addition, when the screen is generated, the skin color measured by the standard face money such as Yu Ping: face P can be displayed only by excluding the skin color of the face. [〇74]又& The difference between the average of each person (age, occupation, gender), the ideal candidate's information, the past data, and the other's qualifications. Therefore, it is helpful to recommend the sale of cosmetics, etc. 19 201028963 [007 5] Next, the details of each main part of the skin color evaluation processing of the above-described embodiment will be described. [0076] <Face division mechanism 24: S02> 5 Next, the face of the above-described face division mechanism 24 The division method is specifically described. The face division unit 24 performs predetermined division on the digital image including the input face. [Fig. 5] Fig. 5 shows an example of the feature point and the division region of the present embodiment. In the present embodiment, for example, as shown in Fig. 5, the entire face of the face is divided into 93 regions, and the average skin color of each of the divided regions is obtained, and the distribution of facial skin color is represented by 93 wide-color data. The skin color evaluation is performed from the distribution or the like. [0078] By taking a lot of information from a certain part of the advantages of the conventional method, Japanese women who can calculate the part (in addition to 'can also be foreigners (the 15 others) , can also be the distribution range or average of the males. As a result, the individual's wide-color data can be compared with the indicators. For example, the skin color comparison before or after the use of the same person's cosmetics can also be performed. The skin color comparison between the two is the same. [0079] The example of the segmentation method shown in Fig. 5 has one to nine feature points. In addition, an example of the segmentation region shown in Fig. 5 is three or four features. 93 areas of a triangle or a quadrangular shape formed by dots (in the fifth figure, the area indicated by numbers 1 to 93). [0080] FIGS. 6A to 6D are diagrams corresponding to the above FIG. An example of the positional relationship in the face of the feature points. The 7Ath to the 7thth figure 20 201028963 does not constitute an example of the combination of the feature points corresponding to the respective regions of the fifth figure. ~ "N〇·", "Name" of each feature point shown in Fig. 6D, and the names of "area N〇·" and "composition point" in each area shown in Fig. 7A to Fig. 7C correspond to the above The content shown in Figure 5. [0081] Here, when the divided content is set, the face dividing means 24 sets the feature points of the front side ν 〇 ι 〜 37 in the feature points shown in FIGS. 6A to 6D (for example, in FIG. 5). In the middle, the point indicated by "Call" is set as the i-th feature point. In addition, these 37 feature points are in the full face area, with 5 points for the forehead part, 10 points for the left and right eyes, 7 points for the nose, 9 points for the mouth, and 6 points for the 10 face lines at the moment. . [0082] Next, the 'face division unit 24' displays the feature points of No. 38 to 109 in the sixth to sixth pictures based on the 37 feature points (first feature points) (in FIG. 5). The point indicated by "△" is set as the second feature point. [0083] For example, as shown in FIG. 6A and the dip figure, the predefined feature points 1 to 37 of n〇1 to 37 are passed through at least two of the feature points and the intersection of the complex lines between the points. The obtained points 38 to 49, the points between the two points are separated by a pre-ratio, 5〇~57, 67~109, and the line passing through the two feature points is 'having the same as a specific point. The ordinates or the coordinates of the ordinates 58 to 66, a total of 1 〇 9 points.

〇 [〇〇84] Further, among the first and second feature points (109 points), at least three of the areas surrounded by the constituent points are as shown in the seventh to seventh embodiments. Further, the number of points constituting the area may be 3 points or 4 points or 5 points or more as shown in Figs. 7A to 7C. Here, each region (region No. 1 to 93) shown in Fig. 5 is set to 21 201028963. Based on the experience of observing many skin colors, it is physiologically divided. In other words, by setting as shown in Fig. 5, the portion which is liable to cause uneven skin color is divided into regions, and the portion where the skin color unevenness is not easily generated is divided into regions. Specifically, in the division example shown in Figs. 5 to 7C, the portion of the forehead is set to be large, and the area around the eyes, the corner of the mouth, or the cheeks is set to be narrow. In this way, the divided area is set to evaluate the important part (area) of the skin color, and the area is set narrowly, so that a highly accurate evaluation can be performed in more detail. ° _ 1〇 [0087] In addition, the 'divided area' determines which skin color is easy to appear in the area based on past data or the like, and collects (groups) based on the degree of the color. By this, it is easy to evaluate each group. [0088] Here, in the above-described fifth to seventh embodiments, the number of feature points defined in advance is 37 (No. 1 to 37), and the present invention is not limited thereto, and at least 15 is set to 25 The feature points (first feature points) can be similarly divided. [0089] Fig. 8 shows an example of another first feature point setting. As shown in Fig. 8 'at least 25 numbers are set! Feature points, so that the & feature points are 4 to 5 points in the forehead part (for example, in Figure 8 +, "Lu" shows 4 points), 8 to 10 points in the vicinity of the left and right eyes (for example, in Figure 8) +, "_" shows 8 points), 5 to 7 points in the nose 2 (for example, in the 8th towel, "♦" shows 5 points), in the mouth is 4:9 (for example, in the 8th) In the figure, '▼' shows 4 points), and the face line at the moment is 4 to 6 points (for example, 4 points in the '+' display in Fig. 8), according to the first ice sign, and the second feature. In the same manner, the same region division as that in the case of using the feature points shown in the above-mentioned Fig. 5 to Fig. 2 can be realized. [0090] 201028963 <Fracture of face and generation of a wide-color distribution map of the face> Split face, specify face In addition, in the present embodiment, the illumination device 5 for uniformly illuminating the entire face and the photographing device using the digital camera are used for the use of the "Family and others, the use of the image analysis speckle, the freckle quantification system" The method of photographing the face image obtained by the development, makeup technique, V28, N2, 1994., etc., and the generation of the skin color distribution circle, the shooting method of the digital image captured by the present invention is not limited to this. 1〇[〇〇91 Here, FIGS. 9A to 9C are diagrams for explaining an example of a flow of face division of an image obtained by photographing. Further, in the examples of FIGS. 9A to 9C, the above-described conventional use is used. The image of the 3rd generation female model A photographed by the photographing device. [0092] First, as shown in FIG. 9A, the '5 test subject set at the predetermined position is photographed, as described above, designated by the face splitting mechanism 24 37 The first feature point 'calculates a total of 109 feature points. [0093] Further, the face division mechanism 24 divides into nineteen feature points from the nine feature points, and is divided into 93 regions by the settings of the seventh to seventh embodiments. When using the average skin color of each region, The result of the application in the area is the image shown in Fig. 9B. 20 In addition, this skin color distribution uses the L*a*b*, Cab*, XYZ color system of the L*a*b* color system for each region. The third stimulus value X, γ, z, RGB values, hue H, brightness V, chroma C, melanin amount, and hemoglobin amount are generated according to the average value. In addition, at this time, L* is used. A*b* color system, parent ¥2 color system, hue Η, lightness v, chroma C, three elements, generate images. 23 201028963 [0〇94] In Figure 9B, due to the external and Not skin color in the area. Htj is not for the evaluation object, so it is colored in a special color such as a color that is far from the skin color. Furthermore, 'in the 9th and 9th pictures, the detailed information of the skin disappears'. The distribution of the complexion of the face is mastered. When the model 5 is shown in Figure 9A, it is known that there is a deep complexion around the eyes. Characteristics. Further, in the skin color evaluation device 2 of the present embodiment, since the uniformity of the illumination of the peripheral portion of the face is low, the skin color distribution evaluation unit 25 can exclude the data of the peripheral portion. Specifically, the skin color distribution evaluation unit 25, in the total of 93 divided regions, as shown in the 9th and 9th, is a predetermined number of frames 41 in the frame 41. The area (61 in Figure 9C of Figure 9b) is used as valid data, and the valid data is used for evaluation processing. * [〇〇96] In addition, in the same way, the "average face" is divided into regions, and each region is colored with 93 skin colors of the model A. As shown in Fig. 9C, 15 can be excluded from the model. Simple skin color information for face shape information.

By performing the above-described processing, it is possible to exclude the face shape information of the person by taking the divided region as a reference, and it is easy to compare the skin color distribution between the face shape $Q and the person. Therefore, it is also possible to use this feature, 20 = the facial skin color distribution of the special A is compared with the average value of the same generation [〇〇98], the above picture and the 9C picture are generated by the image generating means %. The image may also be displayed by the output mechanism 22 to the user or the like, or may be stored in the storage mechanism 23. 24 201028963 <Example of generation of facial skin color distribution for comparison: S04> 5 Next, an example of generation of a facial skin color distribution for comparison of the skin color distribution evaluation unit 25 will be described. Fig. 10A to Fig. 10E show an example of the skin color distribution of the face for comparison. Further, in the examples shown in Figs. 10A to 10E, an example of the skin color distribution of each age is displayed. Fig. 10A to Fig. 10E show the skin color distribution results of the 20th to the 60th generations.

10 15

[0100] As shown in FIGS. 10A to 10E, first, the face image area of the person of the corresponding age is taken out from the storage mechanism 23, and then the LW color system UW, χγζ color system is used. At least one of the stimulus values x, y, z, rGB0, -h, _v, ^Ci, u, and hemoglobin amount, and the skin color distribution of the average value is obtained as the skin color distribution for comparison, and the average value is calculated for each age. Find the average skin color distribution. Further, as shown in Fig. 10A to Fig. 3, each region of the average face is colored using the information of the average skin color distribution of the pure Japanese in each generation of the 20th generation. In addition, the colored image is generated by the screen generating unit 26, and the output unit 22 is rotated or stored in the storage mechanism 23='boring the color display, and the average of the years as shown in the i-th diagram to the first jump diagram can be generated. Skin color distribution. High-accuracy skin color evaluation with this subject image [0102] Data comparison can be performed <Comparison of skin color distribution> Next, in the comparative example of skin color distribution, skin color is used. This embodiment is compared. For example, when the model of the 轴 分布 分布 分布 , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , The upper part of the skin is darker. [0103] Here, a comparative example of the skin color distribution of the difference is taken. In the example shown in Fig. 11, the difference between the amount of the average face and the model A face obtained from the color value 5 of each region is displayed in the frame 41 in a diagonal line. The part with a large amount of melanin. [0104] As shown in Fig. 11, it can be seen that the amount of black pigment in the upper half of the face of the model a is more than the average of the same generation. In addition, the comparison object can be not only the average of the skin color distribution, but also the ideal skin color distribution. 10 [_5] In the present embodiment, by dividing the facial image into predetermined regions, the facial skin color distribution can be grasped, and the skin color can be replaced by the standard face such as the average face, and only the face shape can be excluded. The skin color information is easy to understand. Since the average value of each person in a certain category (age, occupation, gender), the information of ideal candidates such as entertainers, the information of the individual's past 15 and the data of others, etc., Help with selling cosmetics, etc. <Collection of skin color distribution (grouping) and skin color distribution profile generation example> Here, in the present embodiment, the skin color distribution evaluation mechanism 25 will have a similar tendency to skin color in the area of the skin color 2 The area is aggregated (grouped) by analyzing the main components of the past data, such as the main component. In this way, each group can also be evaluated. [0107] Figure 12 shows an example of the area of the grouping. Fig. 13 is a view showing an example of a skin color feature corresponding to each group of Fig. 12 and an area 26 201028963 constituting each group. Further, the feature points constituting the regions shown in Figs. 12 and 13 correspond to the above Figs. 5 to 7 . [0108] In the examples shown in Fig. 12 and Fig. 13, (1) face frequency, (2) cheek front, (3) eye risk, dark eye area, (4) of each main component 5 Forehead, (5) around the nose, (6) group around the mouth. In addition, the skin color feature is "(1) Under the cheek" is high-definition, "(2) The cheek front is reddish and slightly bright, and (3) The dark eye is "yellow" and "low-lightness". (4) Forehead" • It is yellowish and slightly brighter. '(5) Around the nose is reddish and slightly low-light, and "(6) Around the mouth is reddish and slightly low-light. [0109] Here, an example is the effective 57 regions of the four regions excluding the lips of 59 to 67 years old, and the result of performing the principal component analysis of the hue H can be known as 6 principal components. 90.1% of 57 data. [0110] Therefore, in the present embodiment, 57 regions are classified into (1) under the cheek, (2) on the cheek front, (3) on the eyelid, 15 on the black eye, and on the forehead, based on the above-described main components. (5) Around the nose, (6) 6 kinds of around the mouth. Also, # can balance the distribution of the main component scores (skin profile) to assess the distribution of skin color. [0111] Fig. 14 is a diagram showing an example of the evaluation result of the skin color distribution. In the example shown in Fig. 14, the laser map shows the main component scores of the principal components of the above (1) to (6). Line 51 shows an average of 20 points for the main component of the 3rd generation of women, and line 52 shows the main component score of the above model A. [0112] Further, in the example of FIG. 14, an example is obtained by excluding the amount of melanin in 57 regions of the four regions of the lips of the model a, and finding the principal component scores for the six principal components (in FIG. 14). In the middle, the point displayed by "〇" is used to generate a facial skin section map. In the same manner, the average score of the 3〇 generation is calculated as the composition of the composition (in the figure of Fig. 14, ^Λ), and a 3 〇 flat sentence is generated. J surface map. These results are shown in Dinggang. According to this result, it is possible to know at a glance that the skin color distribution characteristics of the individual in the forehead of the model A or the average amount of melanin around the eyes are larger than those of the same generation, and the skin color can be accurately estimated. . In addition, in the above-mentioned _ towel, it is not possible to display six main components, or at least one of them, and the evaluation result is displayed. [0113] Thereby, only six pieces of information evaluation can be provided to the user to provide a Wei index which can grasp the distribution of the skin color at a glance. Further, in the above-mentioned Fig. 14, it is also possible to display not all of the six main components, but also to use at least the species to display the evaluation results.

[0114] In other words, in the skin color evaluation processing program shown in FIG. 4, in the process of S05, the skin color distribution obtained by the process of S03 and the skin color distribution obtained by the process of SO04 are compared, and the skin color is compared. In the present invention, it is not limited thereto, and as shown in FIGS. 12 to 14 , the skin color distribution profile view is generated from the measurement results, and the generated skin color distribution profile view is evaluated in correspondence with the pre-stored comparison skin color distribution profile view. Further, the image or value generated as shown in FIGS. 12 to 14 can be displayed by the output means 22 to the user or the like, or can be stored in the error storing means 23. <Comparison of a conventional method and the present method> Next, a comparison between a conventional method and the present method will be described. Fig. 15 is a diagram showing the difference in hue Η and brightness V of the group, and Fig. 16A shows one of the hues corresponding to the hue of Fig. 15 'the 16th figure is corresponding to Fig. 15 One example of the brightness V. The l6A image shows the horizontal axis 色 28 201028963, and the vertical axis shows the frequency (%). Fig. 16B is a horizontal axis showing the brightness #range' vertical axis display frequency (%). [0117] The present invention performs the skin color evaluation of the digital rib machine, and the money is compared with the chromaticity-like first-line Yang (transfer reading 丨 and #专5利文2) which obtains the data of a certain point, and the subject can be divided in detail. Evaluation of the face is also easy to subdivide as needed. Further, when grouping similar regions of the same color in the face in the face, since the statistical method of principal component analysis is also used, the accuracy is high, and as a result, the non-patent document i will be The face area outside the neck is divided into four groups, and in the present invention, it is divided into six by statistical analysis, and the classification method is also different from the conventional method. Further, regarding the display method of the skin color cross-sectional view, Non-Patent Document 2 displays the actual measured values on the planes of the hue and the lightness, and in the present invention, the radar chart for one color value is used as shown in FIG. , indicating the value of the individual value and the average value of the object as a ratio of 15. In the same way, there is a prior research on the unevenness of the facial skin color (Non-Patent Document 3 and Non-Patent Document 4), which is used to determine the brightness L of all pixels in a certain range. * or the standard deviation of the intensity b* of yellow, or the difference between the L* of the cheek and the chin, does not reach the distribution of the skin color of the entire face 2 of the face. Specifically, in Non-Patent Document 1, the central portion of the forehead ((1)) and (〇2) is classified into a zone A (dark and redish). On the other hand, for the same forehead, the left forehead (〇3) is classified as the area c (dark and ydl〇wish: dark yellow) as the same classification as the left eye (6) 4) and the left eye (〇 5). 29 201028963 [0122] Here, using the present invention, the center portion of the forehead (corresponding to the region > ^.13, 18 in the present invention), the left forehead (area 1^〇.14, 19), and the left eye are investigated. And the color characteristics of the left eye (area No_26, 28, 37), as shown in Figure 15. [0123] That is, as shown in Fig. 15, regarding the hue, the average values of the three regions (the central portion of the forehead, the center of the left, the left and the left, and the left eye) are almost the same, and the histogram shown in Fig. 16 is viewed. There is also the same tendency to obtain the basis for classification into two.

[0124] On the other hand, in terms of brightness ν, since the center of the forehead and the left forehead are similar, the brightness on the left eye and the left eye is low, so the center of the forehead and the left forehead 10 are the same classification, and the upper and lower left eyes should be Different from these classifications. This is also the same as the histogram shown in Figure 16. Thus, in the present invention, according to the above fact, the area corresponding to the central portion of the forehead (〇1) (〇2) Νο·13, 18 and the area corresponding to the left forehead (〇3)> .14,19 is the same classification of "(4) forehead", equivalent to the left eye (〇4)

15 and the left eye (〇5) Ν ο. 26, 28, 37 are different from the "(2) black eye area of the eyelid." Thus, by using the grouping of the face segmentation method of the present invention, skin color evaluation can be performed with greater precision. [0127] As described above, according to the present invention, the skin color can be estimated with a high degree of accuracy of 20 degrees regardless of the shape of the face. Specifically, the input facial image is divided into predetermined regions 'the skin color distribution can be grasped, and the skin color can be replaced by a standard face such as an average face, and the skin color information excluding the face shape can be easily understood. Presented. Moreover, the average value of each person of a certain category or the difference value between the two can be obtained. Further, in the above-described embodiment, the subject of the skin color evaluation is described by the face of 30 201028963, and the present invention is not limited thereto, and may be a wrist or other part of the hand. The above is a detailed description of the preferred embodiments of the present invention, and the present invention is not limited to the specific scope of the invention. 5 [0130] This International Application claims the benefit of priority to the present patent application No. 2008-008370, filed on Jan. 17, 2008, the entire disclosure of which is hereby incorporated by reference. _ [Simple description of the drawing] Fig. 1 shows an example of skin color evaluation by a conventional method. 1〇 Fig. 2 shows an example of the functional configuration of the skin color evaluation device of the present embodiment. The third drawing shows an example of a hardware structure in which the skin color evaluation processing of the present embodiment can be realized. Fig. 4 is a flow chart showing an example 15 of the skin color evaluation processing program of the present embodiment. • $ 5 ® shows one of the feature points and division areas of this embodiment. Fig. 6A shows an example (1) of the positional relationship in the face corresponding to the 1st to 9th feature points of the fifth drawing. 2〇 The _ series shows an example (2) of the positional relationship in the face corresponding to the 1st to 9th feature points in the fifth figure. Fig. 6C shows an example (3) of the positional relationship in the face corresponding to the feature point of the fifth figure. Fig. 6D shows an example (4) of the positional relationship in the face 31 201028963 corresponding to the 109 feature points of the fifth drawing. Fig. 7A shows an example (1) of a combination of feature points constituting each region corresponding to the above fifth graph. Fig. 7B shows an example (2) of a combination of five points constituting the features corresponding to the respective regions of the fifth drawing. Fig. 7C shows an example (3) of a combination of feature points constituting each region corresponding to the above fifth graph. Fig. 8 shows an example of another feature point setting. Fig. 9A is an example (1) for explaining the flow of the face division of the image obtained by the shooting. Fig. 9B is a diagram (2) for explaining a flow of face division of an image obtained by photographing. - Fig. 9C is a diagram for explaining an example of the flow of face division of an image obtained by shooting (3). 15 Fig. 10A shows an example of the skin color distribution for comparison (1). Fig. 10B shows an example of the skin color distribution for comparison (2). 〇 The 10C chart shows one example of the skin color distribution for comparison (3). The 10th figure shows an example of the skin color distribution for comparison (4). Fig. 10E shows an example of the skin color distribution for comparison (5). 20 Figure 11 is a comparison of the skin color distributions for the difference. Figure 12 shows an example of an area classified into groups. Fig. 13 is a view showing an example of the skin color characteristics of the main components of the respective groups in Fig. 12 and the area numbers constituting each group. Fig. 14 is a diagram showing an example of the evaluation result of the skin color distribution. 32 201028963 Figure 15 shows one of the examples of hue and lightness V used to illustrate the difference between groups. Figure 16 shows an example of a histogram corresponding to the hue of Figure 15. Fig. 16 is a diagram showing an example of a histogram corresponding to the brightness V of Fig. 15. [Main component symbol description]

11...average 34... auxiliary memory device 12... 95% trust ellipse 35...memory device 20...skin color assessment device 36...CPU 37...network connection device 21...input mechanism 38. ..recording medium 22...output mechanism 41...box 23...storage mechanism 51,52...line 24...face segmentation mechanism A_ · model 25... skin color distribution evaluation mechanism C... color Degree 26...screen generation mechanism Η...hue 27...control mechanism V...lightness 31...input devices S01 to S09...step 32...output device 33...drive device 33

Claims (1)

201028963 VII. Patent application scope: 1. The method of estimating the skin color from the image of the input face area has the following steps: The correction step is based on pre-sighing the entire face area of the aforementioned image. The i-th feature point formed by at least 25 places and the second feature point set using the ice mark point are divided into predetermined areas; and the skin color distribution evaluation step is performed by using the LW color system in each of the areas divided by the dividing step L*, a*, * 士 lab, Λ 乙 B 10 The skin color distribution of the three stimuli values χ, γ, ζ, dirty values, hue η, lightness V, sensibility C, melanin amount, and hemoglobin amount to m average values, according to the measurement results The evaluation; and the screen generation step, the results of the pre-stab measurement result are displayed on the screen. 2. If applying for the wide-ranging assessment method of the full-time project, the above-mentioned segmentation 15 steps will be at least 25 of the forehead, left and right eyes of the face full bribe Set a number of faces near the eyes, nose, mouth and under the eyes. 3. The skin color evaluation method according to claim 2, wherein the dividing step is divided into 93 regions surrounded by feature points of three or more points selected from the i-th feature point and the second feature point. 2〇4. The skin color evaluation method according to item 1 of the patent application scope, wherein the values of the three stimulus values X, Y, Z, and RGB of the lab*, Cab*, and XYZ color systems using the l*ab color system are At least one of the hue η, the brightness v, the chroma c, the melanin amount, and the hemoglobin amount, the skin color distribution according to the average of the divided regions, and the gradation processing to synthesize the plurality of faces prepared in advance 34 201028963 Partial image ' The evaluation is performed in correspondence with the skin color distribution of each region divided by the average face averaging the face shape. 5 10 15 20 5. For the skin color evaluation method according to item 1 of the patent application scope, the three stimulus values X, γ of the L*a*b*, Cab*, χγζ color system of the L*a*b* color system are obtained. An average value of at least one of Z, RGB values, hue H'lightness v, chroma c, melanin amount, and hemoglobin amount, according to the average value obtained in the similar regions, according to the collected regions The skin color distribution is evaluated. 6. The skin color evaluation method according to item 5 of the patent application, wherein the skin color distribution evaluation step generates a predetermined skin color distribution, a past skin color distribution, a skin color distribution of another person, an average value of a skin color distribution of the plurality, and the foregoing collection. At least one of the skin color distributions of the regions is used as a comparison skin color distribution, and comparison with the personal data to be performed is performed to evaluate the skin color distribution. 7. A skin color evaluation device, which is a person who evaluates a skin color including an input face region, and includes: a 'dividing mechanism, which is based on at least 25 presets of all face regions of the image. The feature point and the second feature point set by the aforementioned Jingzheng point are divided into predetermined regions; the skin color distribution evaluation mechanism is used in each region divided by the dividing mechanism, and the L/a%* color system is used. 1/, a*, yp * a ' b, Cab*, χγΖ The three stimulus values of the color system X, Y, Z, coffee values, hue H, lightness V, chroma c, melanin amount and recorded amount + At least w, the skin color distribution of the mean value of the 'average value' is evaluated according to the measurement result. 35 201028963 The screen generating mechanism displays the aforementioned measurement result or evaluation result on the screen. 8. The skin color evaluation device of claim 7, wherein the segmentation mechanism sets the plurality of face lines of at least 25 of the forehead of the facial image, the vicinity of the left and right eyes, the nose, the mouth, and the underline. 9. The wide color evaluation device according to claim 7, wherein the division mechanism is divided into a plurality of regions surrounded by feature points of three or more points selected from the second feature point and the second feature point. The skin color evaluation device of the seventh aspect of the patent application scope, wherein the skin color distribution evaluation mechanism is a three-stimulus value χ, γ, z of the color system using the L*a*b* color system At least one of RGB values, hue η, brightness V, chroma C, melanin amount, and hemoglobin amount, and the average of each region of the root segment is reddish color distribution, and a plurality of pre-prepared samples are synthesized by gradation processing. The face image is evaluated in correspondence with the skin color distribution of each region divided by the average face averaging the face shape. Η· The skin color evaluation device according to item 7 of the patent application scope, wherein the skin color distribution evaluation mechanism obtains a three-stimulus value of the L*a*b* color system 2L*a*b*, and the XYZ color system, The average value of at least one of γ, z, RGB values, hue Η, monthly V ̄ degree c, melanin amount, and hemoglobin amount is collected according to the average value obtained in the similar regions, according to the collected The skin color distribution of the area is evaluated. 12. The skin color evaluation device of claim n, wherein the skin color distribution evaluation mechanism generates an ideal skin color distribution of the pre-need, a skin color distribution of the past, a skin color distribution of others, and an average skin color distribution of the plurality 36 201028963 And at least one of the skin color distributions of the collected regions is compared with the personal data to be compared as the personal skin material for comparison to evaluate the skin color distribution. 13: A skin color evaluation program, the money includes the image of the face of the person (4), and the image of the person who is the skin of the face is determined by the following steps: the step of dividing is based on the entire face area of the image. The i-th feature point composed of at least 25 pre-sighs and the second feature point set using the first feature point are divided into predetermined regions; and the skin color distribution evaluation step is performed in each of the regions divided by the dividing step. Use the LW color system L*, a*, b*, χγζ The two stimuli values of the color system χ, γ, ζ, RGB values, hue Η, lightness V #> degree 〇 1, melanin amount and blood red At least one of the prime amounts, the skin color distribution of the average value is generated, and the evaluation is performed according to the measurement result; and the step of generating the face is displayed on the face of the measurement result or the evaluation result. 14. A computer readable recording medium recording a skin color evaluation program, the program evaluating a skin color from an image including an input facial region, the recording medium causing a computer to perform the following steps: The first feature point composed of at least 25 positions set in advance on the entire face region of the image and the second feature point set using the first feature point are divided into predetermined regions; the skin color distribution evaluation step is In each of the regions divided by the above-described dividing step, L*a*b* color system L*, a*, b*, cab*, ΧΥΖ colorimetric values of the two stimulus values X, Y, Ζ, RGB are used. , hue η, brightness 37 201028963 V, chroma C, melanin amount and hemoglobin amount at least one, the average skin color distribution is generated, and the evaluation is performed according to the measurement result; and the screen generation step is the measurement result or The evaluation results are shown on the screen. 38