TWI717008B - Method for establishing biometric database, method for facial recognition and system thereof - Google Patents

Abstract

A method for establishing biometric database, a method for facial recognition and a system thereof are provided. In the method, an image reading device is used to capture a plurality of facial images. Multiple environmental light parameters are applied to each of the facial images. A facial recognition technology is incorporated to perform image analysis upon the facial images applied with the environmental light parameters for acquiring biometric features. The environmental light parameters are such as color temperature, color, brightness, contrast and saturation. A biometric database can be established when the multiple biometric features corresponding to the facial images applied with the environmental light parameters are stored. The biometric database is used for facial recognition. For the purpose of facial recognition, the biometric database is used to recognize the user under any of various environmental lights.

Description

Method for establishing biometric database, method and system for face recognition

The present invention relates to a biological feature database, in particular to a method for establishing a biological feature database for performing face recognition under various ambient lights, and a face recognition method and system using the database.

Face recognition technology has been used by various security systems and can be used in applications such as access control, identity recognition, and permission judgment.

However, when performing face recognition, the actual situation of the field may affect the recognition rate of face recognition. The main reason is that when the face recognition database is established, the face image of each person may be taken in a specific field. , Or everyone creates facial images in their respective places, and when performing face recognition, it is performed in another environment, so that the difference in ambient light affects the recognition rate.

In another case, when performing face recognition in a specific field, the field may have changes in ambient light with time or climate. For example, there is a difference in color temperature between morning light and evening light; sunny and cloudy The color temperature of the ambient light in the sky is also different, and it is possible that only a set of fixed or different light source database of facial image characteristics is used, which causes the problem of reduced recognition of face recognition.

According to the method of establishing the biometric database disclosed in the disclosure book and the face recognition method and system using this database, a biometric recognition technology with reference to various environmental light sources is proposed. One of the purposes is to reduce the amount of light The recognition rate is improved.

According to one of the embodiments of the method for establishing a biometric database, the method includes shooting multiple facial images, and applying multiple sets of ambient light parameters to each facial image. The ambient light parameters include color temperature, color, brightness, contrast, and saturation One of the light parameters such as degree, or any combination thereof. After capturing each facial image and applying the biological characteristics of each set of ambient light parameters, after storing the multiple biological characteristics corresponding to each facial image, the multiple biological characteristics formed by applying the multiple sets of ambient light parameters to each facial image are created Biometric database for facial recognition.

Further, when obtaining facial images, the image analysis performed to obtain facial information includes locating the user's facial organs to obtain the position of each organ, and then calculating the contour of each organ, the contour of the face, the proportion of the area, and the number of organs. The ratio of the distance between the two to get the biological characteristics.

In one embodiment, the use of image analysis to obtain face information includes a corresponding high-dimensional space vector generated by a deep network learning model to form biological features for face recognition.

Further, the facial organs used in the biological features in the facial image include pupils, nose tip, mouth shape, chin, earlobe, skin color, eye size, eyebrow length, and eye color, or any combination thereof.

The face recognition method disclosed in the disclosure book uses the biological feature database to perform face recognition. The method first obtains a face image, performs image analysis to obtain the biological characteristics in the face image, and then according to the biological characteristics , After comparing the biometric database, the user can be identified.

Furthermore, the step of analyzing the facial image further includes analyzing the facial image to obtain ambient light parameters, and performing facial recognition based on the biological characteristics and the ambient light parameters recorded in the biometric database to identify the user.

Wherein, in one embodiment, when comparing with a plurality of biological characteristics in the biological characteristic database, the biological characteristics have a corresponding vector, and the difference generated after the comparison of the vectors is lower than a preset value Or when the vectors are closest to each other, it is judged as the same person.

Further, after analyzing the environmental light parameters, the environment classification will be judged first, and when performing face recognition, the biological characteristics in the environment classification will be compared. The basis for judging environmental classification includes the characteristics of light generated by climate, time and occasion.

Further, the face recognition method can be operated in a cloud system with a biometric database, which can provide services for multiple terminal devices to execute the face recognition method through a network.

According to the embodiment of the face recognition system, the main components of the system include an image reading device for obtaining facial images, a device for performing face recognition, and a biometric database that records individual users of multiple users Biometrics of facial images in multiple environmental categories.

According to an embodiment of a face recognition method executed in a face recognition device, a face image is first obtained, and after performing image analysis, the biological characteristics in the face image can be obtained, so that the face can be executed according to the biological characteristics in various ambient light Identify, to identify a user.

In this way, under a concept, the environment where the user is located can be judged and classified according to the ambient light parameters, and these classifications can be based on the light characteristics generated by the climate, time and occasion. When performing face recognition, image analysis technology can analyze the skin color in a specific environment, locate the user’s facial organs, to obtain the position of each organ, and then calculate the contours of each organ, face contour, area ratio and inter-organ The distance ratio is used as the basis for comparing biological characteristics.

Furthermore, the face recognition system further provides a cloud system with the biometric database, and provides services for multiple terminal devices to execute the face recognition method through a network.

In order to further understand the features and technical content of the present invention, please refer to the following detailed description and drawings about the present invention. However, the provided drawings are only for reference and description, and are not used to limit the present invention.

The following are specific examples to illustrate the implementation of the present invention, and those skilled in the art can understand the advantages and effects of the present invention from the content disclosed in this specification. The present invention can be implemented or applied through other different specific embodiments, and various details in this specification can also be modified and changed based on different viewpoints and applications without departing from the concept of the present invention. In addition, the drawings of the present invention are merely schematic illustrations, and are not drawn according to actual dimensions, and are stated in advance. The following embodiments will further describe the related technical content of the present invention in detail, but the disclosed content is not intended to limit the protection scope of the present invention.

It should be understood that although terms such as “first”, “second”, and “third” may be used herein to describe various elements or signals, these elements or signals should not be limited by these terms. These terms are mainly used to distinguish one element from another, or one signal from another signal. In addition, the term "or" used in this document may include any one or a combination of more of the associated listed items depending on the actual situation.

The disclosure book proposes a method for establishing a biometric database, as well as a face recognition method and system using this biometric database. The main concept of establishing this biometric database is to consider the various ambient light conditions when performing facial recognition. The biological characteristics of human faces affected by light. The method is to apply various realistic light sources to the human face when building a face database to create a biological characteristic database with ambient light effects. When performing face recognition under the circumstances, the influence of different light sources can be reduced, thereby increasing the recognition rate.

Refer to FIG. 1 for a schematic diagram of a situation where face recognition is performed under a specific ambient light source.

The figure shows that a user 10 is performing face recognition before the face recognition device 12, and the user 10 is in an environment with a light source 14. The face recognition device 12 uses a camera to capture the face image of the user 10. The image also includes the influence of the light source 14 in the field. At this time, when the face recognition device 12 performs face recognition, it first obtains a face image. When performing face feature analysis, it also analyzes the ambient light characteristics in the face image to obtain the ambient light parameters.

The face recognition device 12 may have a built-in database or be connected to a database in an external device. A biometric database 100 with an ambient light effect for facial images is used. The biometric database 100 records multiple users. Biological features formed by individual users under multiple ambient light sources.

In this way, when the face image is obtained and image analysis is performed, the biological characteristics in the face image can be obtained, and then the ambient light parameters can be obtained from the analysis of the face image, so that the vector characteristics in the biological characteristic database 100, Biometrics or biometrics and ambient light parameters perform facial recognition to identify the user 10.

In one embodiment, when the face image is obtained, the image analysis performed to obtain the face information may include a corresponding high-dimensional space vector generated by a deep network learning model, where the high-dimensional space vector is formed for use by people The biological characteristics of face recognition are used as the basis for identifying the user 10.

According to an embodiment, the ambient light information derived from the face image may include color temperature, color, brightness, contrast, and saturation, and may not Limited to these parameters, one of the parameters or any combination thereof can actually be used, and it can also be equivalent information of these basic parameters, because many designed image parameters are the comprehensive performance of these parameters, such as sharpness and whiteness. Balance (while balance), blur degree (blur) and noise degree (noise degree), etc.

What is mentioned here is that when performing face recognition to identify users, each face image is applied to each set of ambient light parameters to generate multiple sets of corresponding feature vectors to compare with the feature vectors in the biometric database. In one embodiment, if the two feature vectors are compared, and the difference generated after the comparison is lower than a preset value or is the closest vector to each other, it is judged as the same person.

According to another embodiment, when these environmental light parameters are obtained, the environmental classification can be further judged. The basis for judging the environmental classification includes the light characteristics generated by climate, time and occasion. These environmental classifications are the comprehensive performance of these environmental light parameters. For example, according to environmental parameters, the environment in which the user performs facial recognition can be classified into sunlight, cloudy, rainy, etc. according to the climate, and will have different light characteristics; it can be classified according to time, such as morning, noon, evening, and night. All have different light characteristics; they can be classified according to occasions, such as dim yellow light in some occasions, fluorescent lamps in some places, uneven brightness in some places, etc.; can be classified according to photos, such as yellow light, sufficient light, ordinary light, Backlights have different light characteristics. However, the judgment of environmental classification is not a necessary step, but face recognition can be performed through the comparison of feature vectors, biological characteristics, or biological characteristics and environmental light parameters.

When the facial recognition device performs facial recognition to recognize the user, the facial recognition device compares the feature vector, biological feature or biological feature and ambient light in the biological feature database according to the feature vector, biological feature or biological feature and ambient light parameter. The parameters are compared one by one, and face recognition is performed to identify a user. In addition, when the facial recognition device performs facial recognition to recognize the user, the facial recognition device can also determine the environmental classification based on the ambient light parameters extracted from the facial image, and then compare it with the biometric database based on the environmental classification. The feature vector, biological feature, or biological feature of the environment classification is compared with the ambient light parameter, and face recognition is performed to identify a user.

Therefore, once the facial recognition method incorporates the light characteristics, the corresponding biometric database records the biological characteristics of various ambient light effects, which can effectively increase the recognition degree. Moreover, because of the consideration of ambient light, it can first Environmental classification, and the ability to compare biological characteristics under corresponding environmental classifications can improve identification performance.

FIG. 2 shows a diagram of an embodiment of the system for realizing the face recognition method. In the figure, the face recognition technology is described as a plurality of functional modules with software and hardware.

The image reading device 201 is provided with a camera to obtain a face image. According to one of the embodiments, a pre-processing can be performed during the process of obtaining the face image. For example, if necessary, the obtained face can be corrected. The size and angle of the image are stored as image data to facilitate the reliability of subsequent recognition.

The facial recognition device 200 implements facial recognition technology by software or with hardware. For example, in one embodiment, after obtaining a facial image, the biometric capture unit 203 performs image analysis to obtain the facial image. Biological characteristics, the image analysis performed can include skin color analysis, locating the user’s facial organs to obtain the position of each organ, and then calculating the contour of each organ, the contour of the face, the proportion of the area, and the proportion of the distance between the organs. Biological characteristics.

In one embodiment, when the face image is obtained, the image analysis performed to obtain the face information may include a corresponding high-dimensional space vector generated by a deep network learning model, where the high-dimensional space vector is formed for use by people The biological characteristics of face recognition are used as the basis for identifying the user 10.

The environmental feature analysis unit 205 can analyze the light characteristics from the face image, and obtain the ambient light parameters formed by one or any combination of color temperature, color, brightness, contrast, and saturation, and can further obtain the environmental classification to A feature comparison unit 207 is provided to compare the biometrics of each user under the influence of various ambient light in the biometric database 100, and each user is associated with the biometrics and ambient light parameters, so that the facial recognition device 200 The processing procedure can be based on the analyzed environmental parameters or their classification to compare specific environmental parameters or biological characteristics under their classification. It is mentioned here that, in the embodiment, the judgment of environment classification is not a necessary step, but face recognition can be performed through the comparison of feature vectors, biological characteristics, or biological characteristics with ambient light parameters.

The result output unit 209 then outputs the comparison result, which may include a comparison failure, because the biometric database 100 does not record the analyzed biological characteristics under the specific environment classification, or the user can be identified based on the comparison result.

The biometric database 100 described in the above embodiment is a pre-established database, in which a variety of realistic light sources are applied to each face image, and the biometric characteristics of each user under different light sources can be formed to create biometric data. Database, the flowchart of related embodiments can be referred to FIG. 3, and a schematic diagram of establishing a biometric database can be shown with reference to FIG. 4.

In step S301, the registration process is started first, and a device for performing facial recognition or an electronic device equipped with a camera can be used. In step S303, the user's facial image is captured to form a facial image 40 as shown in FIG. 4 , Multiple users generate multiple facial images, and each facial image forms identification information (face ID) when shooting, and correspondingly records the user ID (user ID). In step S305, at this time, the ambient light parameters that have been set under various light sources are applied to the face image to form a realistic light, as shown in Figure 4, where multiple sets of ambient light parameters are applied to the face image 40. The schematic shows that there are a night light effect 401, a sunlight effect 402, and a fluorescent light effect 403. These ambient light effects are respectively composed of various light parameters, and the actual implementation is not limited to these effects.

Each group of ambient light parameters can correspond to at least one environment category and serve as a query index in the biometric database. For example, in a certain climate, a certain time and a certain occasion, after applying various ambient light parameters, in step S307, use Image analysis technology captures the biological characteristics applied with ambient light parameters in each face image to form a variety of biological characteristics as shown in Figure 4, such as applying night light effect 401 to form biological characteristic one 411, and applying sunlight effect 402 to form biological characteristic Two 412, and applying the fluorescent lamp effect 403 to form a biometric feature three 413, after that, in step S309, according to multiple environment classifications, multiple biometric features corresponding to each facial image are stored to form multiple facial images and apply to individual facial images The multiple biological characteristics formed by the above multiple sets of ambient light parameters establish a biological characteristics database 100.

In addition, according to another embodiment, each group of ambient light parameters can correspond to at least one environmental category and serve as a query index in the biometric database. For example, after applying various ambient light parameters in a certain climate, a certain time and a certain occasion , In step S307, use image analysis technology to capture the feature vector applied with ambient light parameters in each face image to form a variety of feature vectors as shown in Fig. 4, such as applying night light effect 401 to capture feature vector 1. Apply The sunlight effect 402 captures feature vector two, and the fluorescent lamp effect 403 captures feature vector three. Then, in step S309, multiple feature vectors corresponding to each facial image are stored according to multiple environmental classifications to form multiple facial images Applying the multiple feature vectors formed by the multiple sets of ambient light parameters to the individual facial images of, the biological feature database 100 is established.

For example, the fields recorded in the biometric database 100 may include user identification data (user ID), including other user information, such as name, gender, age, and job title; then the field records the user's facial image The added ambient light parameters, for example, the R, G, and B values of the night light effect 401 in the red-green-blue (RGB) color space, and the R of the sunlight effect 402 in the red-green-blue (RGB) color space , G, B values, and the R, G, and B values of the fluorescent lamp effect 403 in the red-green-blue (RGB) color space; the next column records other environmental parameters, such as saturation, brightness, contrast, and sharpness , White balance, blur and noise level, etc.; the following fields are the biological features formed between facial organs, which can be a combination of a series of feature vectors, such as "-0.0312066,0.0721339,0.1305,0.0365532,-0.0377493, 0.0769853,-0.127076,0.00105177,-0.0108275,0.0441909,0.0563452,0.0679601,-0.0210059,0.133741,-0.0332901,0.0541074,0.0673971,-0.00912411,0.106956,0.0246078,-0.0929529,-0.0940401,-0.159353,0.0797331,-0.0916072,0.0308622 , -0.0674831,0.130585,0.110465,0.0198231,-0.0272877,-0.0830602,0.0623399,0.110629...etc".

Fig. 5 shows a diagram of another embodiment of a system for implementing a face recognition method. In addition to the application of each local face recognition device, the system for realizing the face recognition method can provide a cloud system 50, and the biometric database 500 can be set in the cloud system 50, so that the cloud system 50 can be connected via the network 52 Provides services for multiple terminal devices 501, 502, 503 to perform facial recognition methods. These terminal devices 501, 502, 503 can be image reading devices installed in various fields, or facial recognition devices with partial functions , And each terminal device 501, 502, 503 does not need a complete biometric database, but is compared through the cloud system 50.

At this time, the biometric database 500 with various ambient light effects set up by the cloud system 50 can meet the needs of image recognition for various settings with different ambient light characteristics in different fields, without the need to perform face registration for individual fields, namely It can achieve high-recognition facial recognition service.

According to the face recognition system described in the above embodiment and the establishment of a biometric database with various ambient light effects, FIG. 6 shows a flowchart of an embodiment of the face recognition method.

In step S601, the face image is obtained by the image reading device. At this time, in the pre-processing, the face image can be corrected, and the color space (such as RGB, YUV, HSV, etc.) and color space processing (such as gray Stepping), skin color analysis, etc., and feature comparison can be used to identify facial organs, locate various facial organs, such as knowing the positions of eyes, nose, ears, lips and eyebrows, and then calculate the contours of each organ , The contour of the face, the proportion of the area, and the proportion of the distance between the organs to obtain the biological characteristics, as in step S603. In another embodiment, the facial organs that can be used for the biological characteristics in the acquired facial image include one of the pupils, nose tip, mouth shape, chin, earlobe, skin color, eye size, eyebrow length, and eye color. , Or any combination of identification.

For example, the biometric value used to recognize a human face can use the distance ratio between organs, such as eye-to-face ratio (such as face width/distance between eyes), eye-mouth ratio (such as distance between eyes/length of mouth), Eye-mouth ratio (such as the distance between the eyes and the mouth/the distance between the two eyes), etc.; moreover, if the light characteristics of the ambient light are added, various image parameters of each organ can be compared, such as the RGB mean of the eyes and the RGB mean of the mouth It is equal to the RGB mean value of the nose.

Then, in step S605, the ambient light parameters can be analyzed from the face image, and in step S607, the environment classification can be determined based on this, as in step S609, when performing face recognition, based on the biometric database established in advance, Comparing the biometric data in a specific environment classification, the comparison process can set a similarity threshold. Only when the similarity exceeds this similarity threshold will it be classified as a conforming biometric. After the comparison is completed, it can be used to identify users ( Step S611).

To sum up, according to the above method of establishing a biometric database and the description of the embodiment of the related face recognition system, the proposed technology will be created after storing the biometrics corresponding to the facial image with multiple sets of ambient light parameters. One of the purposes of the biometric database is to perform face recognition, so that when performing face recognition, the biometric database is used to identify users under the light characteristics of various environments. The comparison conditions include various environmental light source parameters. For example, color temperature, brightness, contrast and saturation, etc., in this way, the use of biological characteristics or biological characteristics and ambient light parameters will improve the recognition rate of the face.

The content disclosed above is only a preferred and feasible embodiment of the present invention, and does not limit the scope of the patent application of the present invention. Therefore, all equivalent technical changes made using the description and schematic content of the present invention are included in the application of the present invention. Within the scope of the patent.

10: User 12: Facial recognition device 14: light source 100: Biometric database 201: Image reading device 200: Facial recognition device 203: Biometrics extraction unit 205: Environmental feature analysis unit 207: Feature comparison unit 209: Result output unit 40: Face image 401: Night light effect 402: Sunlight effect 403: Fluorescent lamp effect 411: Biometric One 412: Biometrics Two 413: Biometrics Three 52: Network 50: Cloud system 500: Biometric Database 501, 502, 503: terminal device Steps S301～S309: the process of establishing a biometric database Steps S601～S611: face recognition process

Figure 1 shows a schematic diagram of a situation in which face recognition is performed under a specific environmental light source;

Figure 2 shows one of the system embodiment diagrams for realizing the face recognition method;

Figure 3 shows a flowchart of an embodiment of establishing a biometric database;

Figure 4 shows a schematic diagram of an embodiment of establishing a biometric database with various ambient light effects;

Figure 5 shows the second embodiment of the system for implementing the face recognition method;

Fig. 6 shows a flowchart of an embodiment of a face recognition method.

40: Face image

401: Night light effect

402: Sunlight effect

403: Fluorescent lamp effect

411: Biometric One

412: Biometrics Two

413: Biometrics Three

100: Biometric database

Claims (18)

A method for establishing a biometric database, including: Shoot multiple facial images; Apply multiple sets of ambient light parameters to each face image; Capture each facial image and apply the biological characteristics of each set of ambient light parameters; and The multiple biometric features corresponding to each facial image are stored to create the biometric feature database in which the multiple biometric features formed by applying the multiple sets of ambient light parameters to each facial image are used for facial recognition. The method for establishing a biometric database according to claim 1, wherein when the facial image is obtained, the image analysis performed to obtain facial information includes locating the facial organs of the user to obtain the positions of the various organs, Then calculate the contour of each organ, the contour of the face, the proportion of the area, and the proportion of the distance between the organs to obtain the biological characteristics. The method for establishing a biometric database according to claim 1, wherein the facial organs used in the biometrics in the facial image include pupils, nose tip, mouth shape, chin, earlobes, skin color, eye size, eyebrow length, and eyes One of the colors, or any combination. The method for creating a biometric database according to claim 1, wherein when the facial image is obtained, the image analysis performed to obtain the facial information includes a corresponding high-dimensional space vector generated by a deep network learning model, The high-dimensional space vector is a biological feature used for face recognition. The method for creating a biometric database according to any one of claims 1 to 4, wherein the groups of ambient light parameters applied to the facial image include: a color temperature parameter, a color parameter, a brightness parameter, and a One of the contrast parameter and a saturation parameter, or any combination thereof. A face recognition method, in which a biometric database is applied, and the method includes: Get a face image; Perform image analysis to obtain biological characteristics in the facial image; According to the biological characteristics, compare the biological characteristics database to perform face recognition to identify a user; Among them, the method of establishing the biometric database includes: Shoot multiple facial images; Apply multiple sets of ambient light parameters to each face image; Capture each facial image and apply the biological characteristics of each set of ambient light parameters; and The multiple biometric features corresponding to each facial image are stored to create the biometric feature database in which the multiple biometric features formed by applying the multiple sets of ambient light parameters to each facial image are used for facial recognition. The face recognition method according to claim 6, wherein, after the creation of the biometric database is completed, when performing face recognition to identify the user, it is generated by applying each set of ambient light parameters to each facial image The corresponding feature vector is compared with the feature vector in the biological feature database, and when the two feature vectors are lower than a preset value or are the closest feature vectors to each other, it is judged as the same person. The face recognition method according to claim 6, wherein analyzing the facial image further includes analyzing the facial image to obtain an ambient light parameter, and executing according to a biological feature recorded in the biological feature database and the ambient light parameter Face recognition to identify the user. The face recognition method according to claim 8, wherein after analyzing the ambient light parameters, an environment classification is determined, and when performing face recognition, the biological characteristics in the environment classification are compared. The face recognition method according to claim 9, wherein the basis for judging the environmental classification includes light characteristics generated by climate, time and occasion. The face recognition method according to any one of claim 6 to 10, wherein the face recognition method operates in a cloud system, and the cloud system is provided with the biometric database, which can provide multiple The terminal device executes the service of the face recognition method. A face recognition system, including: An image reading device for obtaining a face image; A biometric database, which records the biometrics of the facial images of individual users of multiple users under the influence of various ambient lights; A face recognition device for performing image analysis to obtain biological characteristics in the face image, comparing the biological characteristics database according to the biological characteristics, and performing face recognition to identify a user; Among them, the method of establishing the biometric database includes: Shoot multiple facial images; Apply multiple sets of ambient light parameters to each face image; Capture each facial image and apply the biological characteristics of each set of ambient light parameters; and The multiple biometric features corresponding to each facial image are stored to create the biometric feature database in which the multiple biometric features formed by applying the multiple sets of ambient light parameters to each facial image are used for facial recognition. The face recognition system according to claim 12, wherein, after the creation of the biometric database is completed, when performing face recognition to recognize the user, it includes applying various sets of ambient light parameters according to each facial image The corresponding feature vector is generated and compared with the feature vector in the biological feature database. When the two feature vectors are lower than a preset value or are the closest feature vectors to each other, it is judged as the same person. The face recognition system according to claim 12, wherein the facial image obtained by the image reading device is an image under an ambient light, and the ambient light is analyzed to obtain ambient light parameters for Compare the ambient light parameters stored in the biometric database, including one of a color temperature parameter, a color parameter, a brightness parameter, a contrast parameter, and a saturation parameter, or any combination thereof. The face recognition system according to claim 14, wherein in the face recognition device, after analyzing the ambient light parameters, an environment classification is determined to compare the biological characteristics in the environment classification. The face recognition system according to claim 15, wherein, in the face recognition device, the basis for judging the environment classification includes light characteristics generated by climate, time, and occasion. The face recognition system according to claim 12, wherein, in the face recognition device, when obtaining the facial image, the image analysis performed includes locating the facial organs of the user to obtain the positions of the various organs , And then calculate the contour of each organ, the contour of the face, the proportion of the area and the proportion of the distance between the organs to obtain the biological characteristics. The face recognition system according to any one of claims 12 to 17, wherein a cloud system is further provided, provided with the biometric database, and a service for multiple terminal devices to execute the face recognition method is provided through a network .

Images