JP2005352554A - Security system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a security system capable of extracting a same/similar face image even if a face in authentication is varied from that in registration. <P>SOLUTION: A camera device 1 having a face image input part 4 and an authentication server 2 are connected mutually via a communication line network 3. The authentication server 2 is provided with a face image acquisition part 6, an image storage part 11 storing face image data, a registration information generation part 8 cutting a face part around the eyes, the eyebrows, the nose of the face image, calculating a DCT coefficient, generating a binary one-dimensional template, and associating a face image data file name and the binary one-dimensional template by one-to-one correlation, and a database 9 storing a registration template. The authentication server 2 is also provided with an authentication part 7 calculating a DCT coefficient for the face image from the face image input part, comparing the DCT coefficient with the binary one-dimensional template, and extracting one or a plurality of registration templates matching the binary one-dimensional template having correlation above a threshold value and a display part 10 for displaying the registration template. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a security system, and more particularly to a security system that can extract a face image that is the same as or similar to a face image to be authenticated even if the face changes during registration and authentication.

In the conventional face authentication system described in Patent Document 1, a technique is adopted in which authentication is successful when the registered face image and the face image to be authenticated are the same. For this reason, even if the face image of the same person is authenticated, if it gets fat due to aging, the registered face image is different from the face image at the time of authentication, so there is a new problem of inconsistent authentication. there were.
Japanese Patent Application No. 2003-550148

  An object of the present invention is to provide a security system that can extract a face image that is the same as or similar to a face image to be authenticated even if the face changes during registration and authentication.

  The present invention is solved by the following means.

(Claim 1) A camera device having a face image input unit for inputting face image data and an authentication server are connected by a communication line network,
The authentication server includes a face image acquisition unit that acquires a plurality of face image data to be authenticated, an image storage unit that stores the face image data acquired by the acquisition unit, and an eye of the face image extracted from the storage unit Then, the face part in the peripheral region of the eyebrows and nose is cut out, the image data of the cut out face part is DCT transformed, the DCT coefficient is calculated, the binary one-dimensional template is generated, and the face image data to be authenticated is A registration information generating unit that generates a registration template in which an image file name that identifies an image file and the binary one-dimensional template are associated one-to-one, and a database that stores the registration template;
The authentication server includes an authentication face image acquisition unit that acquires photographed face image data to be authenticated input from the face image input unit of the camera device, and eyes, eyebrows, and noses of the face image acquired by the acquisition unit. An authentication information generating unit that cuts out a face part including the image, DCT-converts the image data of the cut out face part, calculates a DCT coefficient, and generates binary one-dimensional information for authentication from the DCT coefficient; and the authentication information generating part The binary one-dimensional information for authentication generated in step 1 is compared with all binary one-dimensional templates stored in the database to determine whether or not the threshold is equal to or greater than a threshold value. An authentication unit that extracts one or more corresponding templates for registration,
The authentication system further includes a display unit for displaying a registration template extracted by the authentication unit.

  (Claim 2) When the authentication server extracts a registration template, the authentication server extracts an image file corresponding to a binary one-dimensional template and displays the image file on the display unit. The security system according to claim 1.

(Claim 3) A camera device including a face image input unit for inputting face image data, an authentication server, and a file server storing a plurality of face image data are connected by a communication network,
The authentication server includes a face image acquisition unit that acquires a plurality of face image data to be authenticated from the file server, a memory that temporarily stores the face image data acquired by the acquisition unit, and a memory that is extracted from the memory A face part of the peripheral area of the eyes, eyebrows, and nose of the face image is cut out, image data of the cut out face part is DCT transformed, a DCT coefficient is calculated, a binary one-dimensional template is generated, and the binary one-dimensional template is generated. Registration for generating a registration template comprising an image file name specified by a one-to-one correspondence between a template and the image file of face image data to be authenticated and the file server name of the file server storing the image file An information generation unit, and a database for storing the registration template,
The authentication server includes an authentication face image acquisition unit that acquires photographed face image data to be authenticated input from the face image input unit of the camera device, and eyes, eyebrows, and noses of the face image acquired by the acquisition unit. An authentication information generating unit that cuts out a face part including the image, DCT-converts the image data of the cut out face part, calculates a DCT coefficient, and generates binary one-dimensional information for authentication from the DCT coefficient; and the authentication information generating part The binary one-dimensional information for authentication generated in step 1 is compared with all binary one-dimensional templates in the database to determine whether or not the threshold is equal to or greater than a threshold value. An authentication unit that extracts a plurality of registration templates,
The authentication system further comprises a display unit for displaying a registration template extracted by the authentication unit.

  According to the present invention, one or a plurality of people can be searched for a registered person, and the searched person is searched for similar persons including the person himself. For example, it is possible to provide a security system that does not miss search even if it changes over time.

  Further, by making the template itself a search item in the DB, data that leaks to the outside, such as an employee number that is originally a search item, is not used, and this has an effect of helping to protect information in the DB.

  Further, when an image file corresponding to the binary one-dimensional template is extracted, the face image can be specifically visually confirmed.

  Hereinafter, embodiments of the present invention will be described.

  FIG. 1 is a functional block diagram showing an example of the security system of the present invention. In FIG. 1, reference numeral 1 denotes a camera device for inputting face image data, 2 denotes an authentication server, and 3 denotes a communication network connecting the camera device 1 and the authentication server 2.

  The camera device 1 includes an image input unit 4. As the image input unit 4, for example, a CCD camera can be used. The camera device 1 also includes a transmission unit 5 that transmits the input face image data to the authentication server 2.

  The authentication server 2 includes at least a face image acquisition unit 6 that acquires a plurality of face image data to be authenticated, an authentication unit 7, an information generation unit 8, a database 9, a display unit 10, and an image storage unit 11. The face image acquisition unit 6 is a face image acquisition unit during registration and an authentication face image acquisition unit during authentication. The information generation unit 8 is a registration information generation unit during registration and an authentication information generation unit during authentication.

(1) Registration Process First, an example of the display unit 10 will be described with reference to FIG. In FIG. 2, reference numeral 100 denotes a registration screen and an authentication screen.

  Reference numeral 101 denotes an item field for display relating to registration processing, a selection button 102 for selecting a place where the acquired face image data exists, a display field 103 for displaying an address indicating the place selected by the selection button 102, and a display field 103. A conversion button 104 is provided for instructing a process of converting each of the face image data at the address displayed in (5) into a registration template and storing it in the database 9.

  Reference numeral 105 denotes an item column for display related to authentication processing, which will be described later. A display column 106 for displaying a place for storing the face image data input by the camera apparatus 1, and a display column 107 for indicating a threshold value for authentication. The authentication button 108 is provided for executing processing for converting the acquired image into binary one-dimensional information and comparing it with the registration template.

  Reference numeral 109 denotes a display field for displaying the corresponding registration template after the authentication button 108 is pressed to perform an authentication process described later, and an authentication rate field for comparing the registration template and binary one-dimensional information created by the authentication process. 110 and a face image file name field 111 registered in the registration template.

  Reference numeral 112 denotes an image display field for displaying an image to be collated. In authentication, an authentication image display field 113 for displaying face image data input by the camera device 1 and a registration template designated in the display field 109 are displayed. A registered image display field 114 for displaying face image data corresponding to any of the registered face image file name fields 111 is provided. Reference numeral 115 denotes a cancel button that is pressed to stop the process.

  Hereinafter, an example of registration processing will be described with reference to FIG.

  First, a registration screen is displayed on the display unit 10 (S1). Next, a selection button 102 is pressed to select a directory (consent with storage location / folder) in which face image data to be acquired is stored. By this operation, the face image acquisition unit 6 acquires a plurality of face image data to be authenticated (S2). The acquired face image data is stored in the storage unit 11. When the selection button 102 is pressed to select a directory, the address of the place where the image displayed in the display field 103 is stored is displayed.

  Here, the face image acquisition unit 6 in the authentication server 2 acquires face image data from a medium such as an optical disk, a magneto-optical disk, or a magnetic disk, or acquires a face image data by photographing a face with the camera device 1. Alternatively, face image data may be acquired through a communication network such as the Internet, an intranet, a VPN, an extranet, or a dedicated line.

  The authentication server 2 may be connected by a communication network such as a dedicated line with a data center that possesses relatively little face image data, and face image data may be received upon request of the authentication server 2. At this time, in the display field 101 of the registration screen in FIG. 2, a button that can be received from the data center storing a plurality of face image data may be provided, or when a selection button 102 is pressed, a plurality of buttons are received. The specification may be such that it can be connected to the data center storing the face image data.

  In the face image acquisition unit 6 that acquires a plurality of face image data to be authenticated, the plurality of face images to be authenticated need only be stored in the storage unit 11 and require a security system. It can also be acquired from a database storing a plurality of face image data.

  The plurality of face image data to be authenticated is basically face image data of a plurality of different persons, but may be a plurality of face image data for the same person. When acquiring a plurality of face image data for the same person, the predicted face image data after shaping may be included.

  When face image data is acquired in the process of S2, the conversion button 104 is pressed. When the conversion button 104 is pressed, the authentication server 2 sends the face image data acquired by the face image acquisition unit 6 to the registration information generation unit 8, and the registration information generation unit 8 performs a binary one-dimensional template creation process. Perform (S3).

  Hereinafter, a system flow of processing for creating a binary one-dimensional template will be described with reference to FIG.

  First, face position detection processing is performed (S20). Specifically, the face position is detected by detecting the pupil from the input face image data.

  Next, the face portion of the area around the eyes, eyebrows, and nose of the face image data is cut out (S21).

  The face part including the eyes, eyebrows and nose of the face image data is cut out because this part is a part with little secular change. FIG. 5 shows a cut-out example, and a portion indicated by a broken line in FIG. 5 is a cut-out portion.

  Next, two-dimensional discrete cosine transform (DCT transform) is performed on the cut face image data (S22).

In the case of input face image information (digital image information) of M × N pixels, when the image signal (pixel value) at the horizontal position m and the vertical position n is f _mn , the horizontal position u on the frequency axis, The DCT coefficient in the vertical direction v is given by F _uv .

The two-dimensional DCT is defined by the following formula. However, the signal f _mn is m = 0... M-1, n = 0.

  Here, C (k) is a normalization constant as shown in the following equation.

Since F _uv is a real number, the separable function of the amplitude part and the sign part (Separable function)

Can be described as However,

It is.

The DCT coefficient is obtained by the above DCT conversion (S23). From the DCT coefficient F _uv , the DCT amplitude and the DCT code (plus and minus) can be obtained.

  The DCT amplitude represents the power at each component (frequency) position. In other words, the DCT amplitude represents the magnitude of the component included in the original signal. The DCT amplitude is information related to the energy of the original signal. It can be said that it bears.

  Even with the same DCT amplitude, if the DCT codes are different, the signals are clearly different. The DCT code has information that is a skeleton of the signal when the signal is configured. That is, the original signal can be identified by the positive / negative arrangement of the DCT code, and the contour line image can be obtained only from the positive / negative DCT code. This becomes apparent when viewing a sign-only synthesis (also referred to as SOS) image, which is an inverse transformation of only a code indicating the characteristic properties of a DCT code.

  Accordingly, the DCT code also reveals the outline of the face of the cutout part and the outlines of the cutout parts of the eyes, nose, eyebrows, and cheeks that are facial parts.

  For example, FIG. 7 shows an example of DCT conversion of 8 × 8 pixel image data shown in FIG. FIG. 7 shows the relationship between DCT coefficients and addresses. FIG. 8 shows a distribution diagram of only code points when DCT coefficients are calculated.

  Next, a predetermined number of DCT coefficient addresses are obtained from the larger absolute value of the DCT coefficient, and a DCT code is extracted from the address (S24). The predetermined number varies depending on the number of pixels, but in the case of 128 × 128 pixels that can be preferably applied in the present invention, about 0.05 to 1% (for example, 8 points to 164) from the larger absolute value of the DCT coefficient. It is a number (number of addresses) determined within a range of about (points). In the illustrated example, there are 8 × 8 pixels, and the predetermined value is 14 points for convenience. That is, the process of S24 is a process of extracting feature points of cut face image data (hereinafter also simply referred to as face image data).

  In order to facilitate understanding of the processing, FIG. 9 shows an extraction point distribution diagram as a result of extracting a predetermined number of addresses from the DCT amplitude. This figure shows the result of extracting 14 points in the case of 8 × 8 pixel image data.

  Next, a binary one-dimensional template represented by binary one-dimensional information “0” and “1” is generated based on the sign of the feature point obtained by the process of S24 (S25). The process of creating this binary one-dimensional template will be described with reference to FIG. That is, when the extraction points in FIG. 9 are superimposed on the code points in FIG. 8, they can be represented as shown in the figure. In FIG. 10, binarized data is acquired for each row in order from the upper left to the right, and binary one-dimensional information is obtained. Based on the result, a binary one-dimensional template as shown in FIG. 11 can be created.

  When the binary one-dimensional template is created as described above, the process returns to the registration process of FIG. 3 and registration is made up of the created binary one-dimensional template and the acquired face image data corresponding to the binary one-dimensional template. A template is created (S4). FIG. 12 shows an example of the created registration template.

  In FIG. 12, 500 is a binary one-dimensional template, 501 is an image file name, and 502 is a registration template comprising a binary one-dimensional template 500 and an image file name 501.

  Here, the image file is a file storing face image data. The name specifying the image file is called an image file name. For example, a state with an extension such as image file name .bmp is called an image file name.

  The registration template 502 is stored in the database 9 of the authentication server 2 (S5).

  FIG. 13 shows an example in which a plurality of registration templates 502 are stored in the database 9 as data tables.

  When the processing up to S5 is completed, the registration processing for one face image data is completed. This is repeated for all of the registered image data to be authenticated, and is performed until the processing of S3 to S5 is completed for all of the registered image data. Here, an example in which the process for one face image data is repeatedly performed is shown, but the processes from S3 to S5 may be performed in parallel with a plurality of face image data.

(2) Authentication process Next, the authentication process will be described.

  FIG. 2 shows an authentication screen. The description of each symbol in FIG. 2 in the authentication process basically uses the one at the time of the registration process.

  Hereinafter, an example of the authentication process will be described based on the process flow of FIG.

  Similar to the registration, the authentication screen 100 is displayed on the display unit 10 (S30). When displayed, the location for storing the authentication face image data is determined. When determined, it is displayed in the display field 106 of the display screen. The storage location may be a temporary storage location or may be stored in the storage unit 11. In the case of temporary storage, it is preferable to provide a memory or the like for temporary storage. The following description will be made in the case of saving in the saving unit 11.

  Next, the authentication button 108 is pressed. When the authentication button 108 is pressed, the camera device 1 is activated to photograph the face of the subject (S31). Next, the authentication face image acquisition unit 6 of the authentication server 2 acquires face image data from the camera device 1 (S32).

  The input face image data is transmitted from the transmission unit 5 of the camera device 1 to the authentication face image acquisition unit 6 of the authentication server 2 and stored in the storage location displayed in the display field 106 of the authentication screen 100. Here, the image name of the captured image is automatically determined as an image file and is stored in the storage location displayed in the display field 106. When saved, the image data is saved as an image file. The image file name has a format of “image file name.extension”. Examples of the extension include JPG, TIFF, GIF, BMP and the like (for example, image file name.bmp, etc.), but are not particularly limited. Further, as the face image file name, the date of shooting, the order of shooting, the device number of the camera, and the like may be configured, and the face image file name is not limited to this.

  Next, the authentication face image acquisition unit 6 sends the face image data to the authentication information generation unit 8, and the authentication information generation unit 8 performs face position detection processing (S33). Specifically, when a face image is input, a pupil is detected from the image.

  The face portion including the eyes, eyebrows and nose of the face image detected in the process of S33 is cut out (S34). The face part including the eyes, eyebrows and nose of the face image is cut out because this part is a part with little secular change. The cutout portion is shown in FIG. 5 as in the registration process.

  Next, DCT conversion is performed on the cut-out face image data by the same method as at the time of registration (S35). Next, DCT coefficient calculation is performed (S36). FIG. 15 shows a distribution diagram of only code points when DCT coefficients are calculated.

  Next, DCT codes corresponding to the DCT coefficients up to a predetermined number from the upper absolute value of the DCT coefficients are obtained (S37). This process is the same as the registration process, and a predetermined number of definitions are the same as the registration process. The process of S37 is a process of extracting feature points of the face image.

  In order to make it easy to understand the process of S37, FIG. 16 shows an extracted point distribution diagram as a result of extracting the feature points of the face image. This figure shows the result of extracting 14 points in the case of an 8 × 8 pixel image.

  Authentication binary one-dimensional information represented by binary one-dimensional information “0” and “1” is created based on the sign of the feature point (S38). A process of creating the binary one-dimensional information for authentication will be described with reference to FIG. That is, when the extraction points in FIG. 16 are superimposed on the code points in FIG. 15, they can be represented as shown in FIG. In FIG. 17, binarized data is obtained for each row in order from the upper left to the right, and binary one-dimensional information is obtained. Based on the result, the binary one-dimensional information for authentication as shown in the upper part of FIG. 18 can be generated. The authentication information generation unit 8 sends the binary one-dimensional information for authentication to the authentication unit 7, and the authentication unit 7 acquires the binary one-dimensional information for authentication.

  Next, the authentication unit 7 searches the binary one-dimensional template in the database 9 based on the acquired binary one-dimensional information for authentication. An example of the search is shown in FIG. That is, the binary one-dimensional information for authentication is compared with the binary one-dimensional template to determine whether the authentication rate is equal to or higher than the threshold (S39), and the image corresponding to the binary one-dimensional template that matches the threshold or higher. A file name is extracted (S40). The processes of S39 and S40 are repeated until all binary one-dimensional templates stored in the database 9 are compared.

  Specifically, there are 14-bit registration information and authentication information. For each bit, the binary one-dimensional template at registration and the binary one-dimensional information for authentication are compared. The numerical value of each bit is calculated by a logical operation called exclusive OR operation, and it is performed for all the bits. When 0 is output, the number is counted, and the number obtained by counting 0 is the number of hits. The authentication rate can be determined by how many hits the number of hits matches. For example, in the case of 11 points out of 14 bits, the authentication rate is about 79%.

  Here, the exclusive logical sum is 0 if each bit of the binary one-dimensional template at the time of registration and the binary one-dimensional information for authentication at the time of authentication is the same value, and 1 if it is different. It is a logical operation. For example, when an exclusive OR operation is performed on 11110001 and 10010011, the result is 01100010, and five of the eight points match, so the authentication rate is 62.5%. When the threshold is set to 80%, the image file name is extracted from the registration template in which the binary one-dimensional template matches 12 points or more in 14 bits.

  As a result of the authentication in S40, the corresponding item is sent to the display unit 10 of the authentication server 2 (S41) and displayed in the registration template display field 109 of the authentication screen 100. FIG. 19 shows an example of a display screen from which three registration templates are extracted from the registration templates.

  By selecting an image file name displayed in the registration template display field 109 of the authentication screen 100, an image file corresponding to the image file name existing in the image storage unit 11 is registered in the registered image display field of the image display field 112. 114. Further, the face image taken at the time of authentication is displayed in the authentication image display field 113. Thereby, the face image data can be visually observed at the time of registration and at the time of authentication.

  The preferred embodiment of the present invention has been described above. However, the embodiment of the present invention is not limited to the above. For example, when the number of extractions is large in the process from S39 to S41, for example, about 20 people are extracted. If it has been done, only those higher than the average value of the extracted authentication rates may be displayed. Alternatively, the threshold value may be increased by about 5% and the authentication process may be performed again.

  Further, the processes from S39 to S41 are processes performed on one registration template, and it is preferable to repeat the processes of S39 and S40 in order to realize the present invention. Alternatively, one registration template may be processed, and the corresponding registration template may be sent line by line to the display unit 10 and displayed in the registration template display field 109. That is, you may repeat the process of S39, S40, and S41.

  In the registration template display field 109, 110 is an authentication rate display field, and 111 is an image file name display field. In this case, the authentication rate display column 110 displays the authentication rate, but the person name of the file name may be displayed.

  The registration template created during the registration process has a one-to-one correspondence between the binary one-dimensional template and the face image. The face image is stored separately, and the registration template contains the face image. Instead, the image file name may be used, and the image file name and the face image may be associated with each other one to one.

  Further, the registration template is composed of at least a binary one-dimensional template, an image file or an image file name, and more preferably, an image person name, an attribute, whether or not to arrange for appointment, and attribute information for an image person such as an offense May also be included. Also, when extracting, you can select the item you want to display before processing from the information configured in the registration template, and extract data that matches the item you want to display in the registration template. it can.

  Furthermore, it is preferable that the security system of the present invention adopts the aspect shown in FIG. This aspect is an example in which the file server 400 is provided separately from the authentication server 2.

  If there is a huge amount of face image data to be authenticated, it is not preferable to import all of the image data into the authentication server 2 as shown in FIG. Therefore, the file server 400 is provided separately from the authentication server, the storage unit 11 of the authentication server 2 is not provided, and instead, a memory 401 that temporarily stores image data is provided, and face image data is stored in the memory 401, After performing the registration process of the present invention, the face image data in the memory 401 may be deleted.

Block diagram showing an example of a security system according to the present invention The figure which shows the display screen at the time of registration and authentication of this invention Process flow diagram showing an example of registration processing of the present invention Process flow diagram showing binary one-dimensional template creation processing of the present invention The figure which shows the example of the cutout part of this invention The figure which shows the image of 8x8 pixel of this invention DCT coefficient distribution diagram after DCT conversion of an 8 × 8 pixel image at the time of registration in the present invention Distribution diagram showing DCT code at the time of registration in the present invention Distribution diagram showing a predetermined number of addresses from the top as absolute values of DCT coefficients at the time of registration in the present invention The figure which superposed | stacked FIG. 8 and FIG. 9 of this invention The figure which shows the example of the binary one-dimensional template of this invention The figure which shows the example of the template for registration of this invention The figure which shows the example of the data table of the database in this invention Process flow diagram showing an example of authentication processing of the present invention Distribution diagram showing DCT code at the time of authentication in the present invention Distribution diagram showing a predetermined number of addresses from the top in absolute value of DCT coefficient at the time of authentication in the present invention FIG. 15 and FIG. 16 are superimposed views of the present invention. The figure which shows the example which contrasted the binary one-dimensional template and binary one-dimensional information in the case of the authentication process in this invention. The figure which shows the example of the display screen displayed as a result of performing the authentication process of this invention The block diagram which shows an example of the security system which concerns on the other aspect of this invention.

Explanation of symbols

1: Camera device 2: Authentication server 3: Communication network 4: Input unit 5: Transmission unit 6: Face image acquisition unit 7: Authentication unit 8: Information generation unit 9: Database 10: Display unit 11: Storage unit 400: File Server 401: Memory

Claims (3)

A camera device having a face image input unit for inputting face image data and an authentication server are connected by a communication network,
The authentication server includes a face image acquisition unit that acquires a plurality of face image data to be authenticated, an image storage unit that stores the face image data acquired by the acquisition unit, and an eye of the face image extracted from the storage unit Then, the face part in the peripheral region of the eyebrows and nose is cut out, the image data of the cut out face part is DCT transformed, the DCT coefficient is calculated, the binary one-dimensional template is generated, and the face image data to be authenticated is A registration information generating unit that generates a registration template in which an image file name that identifies an image file and the binary one-dimensional template are associated one-to-one, and a database that stores the registration template;
The authentication server includes an authentication face image acquisition unit that acquires photographed face image data to be authenticated input from the face image input unit of the camera device, and eyes, eyebrows, and noses of the face image acquired by the acquisition unit. An authentication information generating unit that cuts out a face part including the image, DCT-converts the image data of the cut out face part, calculates a DCT coefficient, and generates binary one-dimensional information for authentication from the DCT coefficient; and the authentication information generating part The binary one-dimensional information for authentication generated in step 1 is compared with all binary one-dimensional templates stored in the database to determine whether or not the threshold is equal to or greater than a threshold value. An authentication unit that extracts one or more corresponding templates for registration,
The authentication system further comprises a display unit for displaying a registration template extracted by the authentication unit.   The authentication server extracts an image file corresponding to a binary one-dimensional template when the authentication unit extracts a registration template, and displays the image file on the display unit. The security system described. A camera device having a face image input unit for inputting face image data, an authentication server, and a file server storing a plurality of face image data are connected by a communication network,
The authentication server includes a face image acquisition unit that acquires a plurality of face image data to be authenticated from the file server, a memory that temporarily stores the face image data acquired by the acquisition unit, and a memory that is extracted from the memory A face part of the peripheral area of the eyes, eyebrows, and nose of the face image is cut out, image data of the cut out face part is DCT transformed, a DCT coefficient is calculated, a binary one-dimensional template is generated, and the binary one-dimensional template is generated. Registration for generating a registration template comprising an image file name specified by a one-to-one correspondence between a template and the image file of face image data to be authenticated and the file server name of the file server storing the image file An information generation unit, and a database for storing the registration template,
The authentication server includes an authentication face image acquisition unit that acquires photographed face image data to be authenticated input from the face image input unit of the camera device, and eyes, eyebrows, and noses of the face image acquired by the acquisition unit. An authentication information generating unit that cuts out a face part including the image, DCT-converts the image data of the cut out face part, calculates a DCT coefficient, and generates binary one-dimensional information for authentication from the DCT coefficient; and the authentication information generating part The binary one-dimensional information for authentication generated in step 1 is compared with all binary one-dimensional templates in the database to determine whether or not the threshold is equal to or greater than a threshold value. An authentication unit that extracts a plurality of registration templates,
The authentication system further comprises a display unit for displaying a registration template extracted by the authentication unit.