JP2002329204A - Personal authentication system and personal authentication program - Google Patents

Abstract

(57) [Summary] [Problem] To shorten the time required for iris code collation processing for personal authentication and improve the processing capability. SOLUTION: An iris code of a person to be authenticated is created by a matching terminal 4, and the iris code is compared with feature data registered in a database of a registration terminal 2 to perform personal authentication. The iris code having the highest matching rate among the feature data having the matching rate equal to or higher than the value is used as the matching matching data, and the number of matching matching times of the iris code is updated. Is deleted from the database, thereby reducing the number of iris codes used for collation.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a personal authentication system and a program for use in various security systems and the like, and more particularly to a technique for managing characteristic data in personal authentication using characteristic data obtained from a human body. It is.

[0002]

2. Description of the Related Art As a system for performing personal authentication using characteristic data obtained from a human body, there is a system using an iris (iris) or a facial appearance. For example, a personal authentication system using an iris captures an eye of a registrant with a camera, creates an iris code (iris pattern data) as feature data from the image, registers the iris code in a database (hereinafter, DB), Then, at the time of personal authentication, the eye of the person to be authenticated is photographed with a camera, and the iris code created from the image is compared with the iris code registered in the DB to determine whether or not the person is a registrant. It is a system to authenticate.

In a conventional personal authentication system, the same person often registers a plurality of iris codes of the same eye in order to improve the accuracy of iris verification. on the other hand,
The matching method when matching the iris code registered in the database with the iris code of the person to be authenticated can be performed by matching all the iris codes registered in the DB and specifying the user ID. There is a designated collation for retrieving an iris code specified by the user ID and collating the retrieved iris code. Which collation method is adopted is determined by the number of registrants, operating conditions, and the like. .

[0004]

However, the above-mentioned prior art has the following problems. In other words, in this type of system, as described above, one person has registered a plurality of iris codes, that is, feature data of the same eye, but among them, there is feature data with a low coincidence rate and a low matching frequency. In the case of a search, all registered feature data, including feature data with low matching frequency, must be compared and matched with the feature data of the person to be authenticated. There is a problem that it takes a long time.

[0005] In the case of designated collation in which characteristic data specified by a user ID is searched and the retrieved characteristic data is collated, less characteristic data is used for collation than in the case of a full search. There is a problem that if there is feature data with a low frequency, the time required for matching becomes longer. An object of the present invention is to solve such a problem.

[0006]

Therefore, according to the present invention, feature data created from an image of a specific part of a registrant is registered in a database, and the characteristic data is created from an image of a person to be authenticated. In a personal authentication system for performing personal authentication by comparing feature data registered in the database with feature data registered in the database, a plurality of feature data of the same registrant can be registered in the database, and feature data of all registrants can be registered. The number of times of matching and matching is set for each feature, and the matching of feature data at the time of personal authentication is performed using the feature data with the highest matching rate among feature data with matching rates equal to or higher than the reference value as matching matching data. Updating the number of times, and deleting from the database feature data that is determined to have a low matching frequency from the number of matching matches. That.

[0007]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, an embodiment of a personal authentication system according to the present invention will be described with reference to the drawings. First, a first embodiment will be described with reference to FIGS. FIG. 1 is a block diagram showing a system configuration according to an embodiment. In FIG. 1, reference numeral 1 denotes a camera for photographing an eye of a registrant;
Is a registration terminal having a database (hereinafter, DB) not shown, 3 is a camera for photographing the eye of the person to be authenticated, and 4 is a collation terminal.

Here, the camera 1 is connected to the registration terminal 2 and the camera 3 is connected to the verification terminal 4, respectively. The registration terminal 2 and the verification terminal 4 are connected by a communication line 5 such as a LAN. In the system having such a configuration, the camera 1 photographs the registrant's eye with the camera, and the registration terminal 2 creates an iris code, which is feature data, from an image obtained by photographing the camera 1 and stores the iris code in the database DB. It has a function to register.

Further, at the time of iris code collation for personal authentication, the camera 3 takes an image of the eye of the person to be authenticated, and the collation terminal 4 creates an iris code from an image obtained by photographing the camera 3 to create the iris code. Has a function of collating with the iris code registered in the. FIG. 2 is a diagram showing an example of registered contents of the DB according to the first embodiment. As shown in FIG. 2, the DB provided in the registration terminal 2 includes a user information DB, an iris code DB, and a collation processing statistical DB.
It consists of.

[0010] Here, the user information DB registers and registers the user ID, name, telephone number, and access rank of the registrant. The iris code DB stores the user ID, iris code number of the registrant. , An iris code, and an eye indicating whether the eye is a right eye or a left eye are registered in association with each other. Further, the user ID of the registrant, the iris code number, and the number of matching matches are registered in the collation statistical database. ing.

As can be seen from the above, the user ID is unique to the registrant, and is common to the user information DB, the iris code DB, and the collation statistics DB, and the registrant having the same iris code number has the iris code number. It represents the order of registration when a plurality of codes are registered, and the iris code of each DB, the number of matching matches, and the like can be specified by a combination of a user ID and an iris code number.

Next, the operation of the first embodiment will be described with reference to all-item matching as an example. First, when the iris code is registered by the registration terminal 2, the user information DB of FIG.
Then, the contents of the iris code DB are added and updated. However, since the same person may perform a plurality of registrations, in such a case, only the iris code DB is added or updated.

The collation terminal 4 performs collation processing using the DB of the registration terminal 2 and updates the collation processing statistical DB of FIG. 2 based on the collation result. The updating of the collation processing statistical DB will be described in more detail. When performing personal authentication,
The camera 3 captures an image of the eye of the person to be authenticated, and the verification terminal 4 creates an iris code from an image obtained by capturing the image of the camera 3.

The verification terminal 4 requests the registration terminal 2 for the iris code registered in the DB. Accordingly, the registration terminal 2 sequentially calls all the iris codes registered in the iris code DB and sends them together with the user ID and the iris code number to the matching terminal 4. The created iris code is collated with the iris code sent from the registration terminal 2.

This collation is performed using all the iris codes registered in the iris code DB. If there is an iris code whose coincidence rate is higher than a predetermined reference value, the person to be authenticated is the same as the registrant. The user is authenticated as a person, and the user ID and the iris code number corresponding to the iris code are returned to the registration terminal 2 as the best matching information, but if there is a plurality of iris codes with a matching rate equal to or higher than a predetermined reference value, The user ID and the iris code number corresponding to the iris code having the highest matching rate are returned to the registration terminal 2 as the best matching information.

The registration terminal 2 searches the collation processing statistical DB for information on the number of collation matching times corresponding to the user ID and the iris code number returned from the collating terminal 4, and performs "+1" processing on the number of collation matching times. To update information. By such processing, the content of the collation statistics DB is updated at any time. Therefore, the registration terminal 2 performs the processing of the collation statistics DB in the collation statistics DB as a process for each fixed period such as daily or monthly. Is used to delete unnecessary information, thereby slimming down the DB.

As a specific process, first, a user ID and an iris code number having a plurality of user IDs and eye information are searched from the iris code DB, and the searched user ID and iris code number are compared with a matching process. By referring to the number of matching matches in the statistical DB, for example, the number of iris codes less than 10% of a certain number of matching matches is deleted.

For example, in FIG. 2, the user ID "6"
Since there are two “right eyes” of “3037” and two “left eyes” of user ID “12077”, both of them are processed. Then, as for the iris code numbers “1” and “2” of the user ID “63037”, the user ID “6”
Since the number of matching matches for the iris code number “1” of “3037” is 12 and the number of matching matches for the iris code number “2” is 1, for example, the two matching matching times of “1”
"2" and "1" are added, and "13" of the addition result is set as the reference number.

In this case, since the number of matching matches of the iris code number “1” is 12, the reference number of times “13” is 1
However, since the number of matching matches of the iris code number "2" is one, it is less than 10% of the reference number "13". Therefore, the iris code corresponding to the iris code number "2" is changed. delete. In addition, the user ID “1”
For the iris code numbers “2” and “3” of “2077”, the number of matching matches of the iris code number “2” is 11
Since the number of times of matching of the iris code number “3” is nine, the reference number is “20”, and the number of times of matching and matching of the iris code number “3” and the number of matching matches of the iris code number “2” are both standard. Since the number of times is equal to or more than 10% of “20”, both are determined to be valid, and the deletion processing is not performed.

FIG. 3 is a diagram showing an example of the registered contents of the DB after such deletion processing. According to FIG. 3, the second user ID, the iris code number, and the iris code from the top of the iris code DB in FIG. It can be seen that the code, the eye, the second user ID, the iris code number, and the number of matching matches from the top of the matching statistical database have been deleted. According to the first embodiment described above, among the iris codes registered in the DB, the iris code having a low matching frequency in actual collation is deleted, so that the iris code used for collation can be reduced. As a result, the processing time can be shortened accordingly, and the effect that the processing capacity can be improved can be obtained.

Next, a description will be given of a second embodiment in which a user ID is input and personal verification is performed on the information.
The configuration of the system is the same as in the first embodiment. FIG.
Is a diagram showing an example of registered contents of a DB according to the second embodiment. As shown in the figure, in this embodiment, information of the number of times of non-conformity of collation is stored in the collation statistics DB for the registered contents of FIG. It has been added.

Other contents in the collation statistical database,
FIG. 2 shows the contents of the user information DB and the iris code DB.
Is the same as Next, the operation of the second embodiment will be described with reference to an example of designated collation for inputting a user ID. First, when the iris code is registered in the registration terminal 2, the user information DB of FIG.
Is added and updated.

However, since the same person may make a plurality of registrations, in such a case, only the iris code DB is added or updated. The verification terminal 4 is a DB of the registration terminal 2
And performs a collation process, and updates the collation statistics DB of FIG. 4 based on the collation result. This matching processing statistics D
The updating of B will be described in more detail.

When performing personal authentication, the camera 3 takes an image of the eye of the person to be authenticated, the collating terminal 4 creates an iris code from an image obtained by photographing the camera 3, and the person to be authenticated performs the collation. The user inputs his / her user ID by using a keyboard (not shown) provided in the terminal 4. Thereby, the verification terminal 4 sends the user ID input by the person to be authenticated to the registration terminal 2 and requests the iris code registered in the DB.

The registration terminal 2 searches the iris code corresponding to the received user ID from the iris codes registered in the iris code DB, and sends the searched iris code to the matching terminal 4 together with the user ID and the iris code number. Then, the verification terminal 4 receives the iris code created from the image of the person to be authenticated,
Check with the iris code sent from.

If there is an iris code whose coincidence rate is equal to or greater than a predetermined reference value, the person to be authenticated is authenticated to be the same person as the registrant, and the user ID and iris code number corresponding to the iris code are entered. The matching information is returned to the registration terminal 2 as the best matching information. If there are a plurality of iris codes having a matching rate equal to or more than a predetermined reference value, the user ID and the iris code number corresponding to the iris code having the highest matching rate are best matched. The information is returned to the registration terminal 2 as information.

The registration terminal 2 searches the collation processing statistics DB for information on the number of collation matching times corresponding to the user ID and the iris code number returned from the collating terminal 4, and performs "+1" processing on the number of matching collation times. To update information. Further, the user ID corresponding to the other iris code sent to the matching terminal 4 and the information on the number of times of non-matching corresponding to the iris code number are searched from the matching process statistical DB, and "+1" processing is performed on the number of times of non-matching. To update information.

With such processing, the collation processing statistical DB
Is updated at any time, the registration terminal 2 uses the information of the number of matching matches and the number of non-matching matches in the matching statistics DB as processing for each fixed period such as daily or monthly. A process for deleting information is performed to slim down the DB. As the specific processing, first, a user ID and an iris code number having a plurality of pieces of user ID and eye information are searched from the iris code DB, and the searched user ID and the iris code number are searched for in the collation processing statistics DB. By referring to the information on the number of times of collation matching and the number of times of non-matching, for example, the iris code of the number of times of matching and matching ≦ the number of times of non-matching is deleted. However, if the same Eye of the user ID after the deletion becomes zero, it is not deleted.

For example, in FIG. 4, the user ID "6"
Since there are two “right eyes” of “3037” and two “left eyes” of user ID “12077”, both of them are processed. Here, as for the iris code number “1” of the user ID “63037”, the number of times of matching is eight times and the number of times of non-matching is one. "6303
For the iris code number “2” of “7”, the number of matching matches is 6 and the number of matching mismatches is 3, so the relationship between the two is “matching matching number> matching mismatch number”.

Therefore, the iris codes corresponding to the iris code numbers "1" and "2" of the user ID "63037" are not deleted. On the other hand, the user ID "120
As for the iris code number “2” of “77”, the number of matching matches is 6 and the number of matching mismatches is 0, so the relationship between the two is the number of matching matches> the number of matching mismatches. Therefore, the iris of this user ID “12077” Although the iris code corresponding to the code number “2” is not deleted, the iris code number “3” of the user ID “12077” has three times of matching and non-matching, and the relationship between the two is matched. The number of matching times is equal to the number of non-matching times. Therefore, the iris code corresponding to the iris code number “3” of the user ID “12077” is deleted.

FIG. 5 is a diagram showing an example of the registered contents of the DB after such deletion processing. According to FIG. 5, the sixth user ID, the iris code number, and the iris code from the top of the iris code DB in FIG. It can be seen that the code, the eye, the sixth user ID, the iris code number, and the number of matching matches from the top of the matching statistical database have been deleted. According to the second embodiment described above, similar to the first embodiment, an iris code used for collation is deleted in order to delete an iris code registered but having a low frequency of matching in actual collation. , The processing time can be shortened accordingly, and the effect that the processing capacity can be improved can be obtained.

In the above-described first and second embodiments, the iris code which has a low matching frequency in the collation at the time of personal authentication is deleted. However, the iris code DB is replaced with the iris code primary DB and the iris code. It is also possible to improve the processing capability by setting the priority of the search order of the iris code by dividing into two code secondary DBs.

FIGS. 6 and 7 show examples of the use of DB registration in this case. Here, FIG. 6 corresponds to the first embodiment,
FIG. 7 corresponds to the second embodiment, in which iris codes that match less frequently in matching at the time of personal authentication are registered in the divided iris code secondary DB. In this way, at the time of the next personal authentication, collation is performed using the iris code of the iris code primary DB, and for the iris code that has become incompatible with the collation, the iris code is registered in the iris code secondary DB. Iris code can be used.

In this case, the iris code secondary DB
If the iris code registered in the iris code is used, it takes a long processing time. However, by performing the matching using the iris code of the iris code primary DB having a high probability of matching first, The time for the matching process is shortened, and as a result, the processing capability is improved. Although the embodiments have been described above, the present invention is not limited to these embodiments.

For example, in the above-described embodiment, the iris code is transmitted from the registration terminal 2 to the collation terminal 4 at the time of collation for personal authentication, and the iris code is collated by the collation terminal 4. The iris code created in step (1) may be sent to the registration terminal 2 so that the registration terminal 2 checks the iris code. Further, in the above-described embodiment, the iris code is described as an example of the feature data for performing personal authentication. However, a face of a registrant is photographed with a camera, and a face code is created from the image. Is also applicable to a system for performing personal authentication using the feature data as the feature data.

[0036]

As described above, according to the present invention, among the registered personal authentication feature data registered in the database, feature data that is not frequently matched in actual matching is deleted. Since the feature data used for the collation can be reduced and the processing time can be shortened accordingly, the effect that the processing capability can be improved can be obtained.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a system configuration according to an embodiment.

FIG. 2 is a view showing an example of registered contents of a DB according to the first embodiment;

FIG. 3 is a view showing an example of registered contents of a DB after a deletion process according to the first embodiment;

FIG. 4 is a diagram showing an example of registered contents of a DB according to the second embodiment.

FIG. 5 is a diagram illustrating an example of registered contents of a DB after a deletion process according to the second embodiment;

FIG. 6 is a view showing an example of registered contents when an iris code DB is divided.

FIG. 7 is a diagram showing an example of registered contents when an iris code DB is divided.

[Explanation of symbols]

 1 camera 2 registration terminal 3 camera 4 verification terminal

 ────────────────────────────────────────────────── ─── Continuing on the front page (72) Inventor Takahiro Tachiki 1-12-11, Funatari, Itabashi-ku, Tokyo Oki Software Co., Ltd. F-term (reference) 5B043 AA01 AA09 BA04 CA09 DA05 FA02 FA07 FA08 FA10 HA02 5B047 AA23 BB06 5B085 AE02 AE23 AE25

Claims (8)

[Claims] 1. A feature data created from an image image of a specific part of a registrant is registered in a database, and the feature data created from the image image of the person to be authenticated is registered in the database. In a personal authentication system that performs personal authentication by comparing feature data, a plurality of feature data of the same registrant can be registered in the database, and the number of matching matches is set for each feature data of all registrants. At the time of personal authentication, feature data with the highest matching rate among feature data with a matching rate equal to or higher than the reference value is used as matching matching data, and the number of matching matching times of the feature data is updated. A personal authentication system, wherein feature data determined to have a low matching frequency is deleted from the database. 2. Feature data created from an image image of a specific part of a registrant is registered in a database, and feature data created from an image image of a person to be authenticated is registered in the database. In a personal authentication system that performs personal authentication by comparing feature data, a plurality of feature data of the same registrant can be registered in the database, and the number of matching matches and the number of matching matches are set for each feature data of all registrants. Is set at the time of personal authentication, the feature data having the highest matching rate among the feature data having a matching rate equal to or higher than the reference value is set as matching matching data, and the number of matching matching times of the feature data is updated. , The feature data other than the verification conformance data is regarded as the verification non-conformance data, and the number of verification failures of the characteristic data is updated. And a feature authentication system that deletes feature data determined as having a low matching frequency by referring to the number of times of non-matching with the database from the database. 3. A feature data created from an image image of a specific part of the registrant is registered in a database, and the feature data created from the image image of the person to be authenticated is registered in the database. In a personal authentication system for performing personal authentication by collating feature data, a primary database and a secondary database are provided in the database, all feature data are registered in the primary database, and a plurality of feature data of the same registrant are stored in the database. In addition to being able to register in the database, the number of matching matches is set for each registrant's feature data. Using the high feature data as the matching matching data, updating the matching matching count of the feature data, A personal authentication system characterized in that feature data determined as having a low matching frequency based on the number of matching matches is deleted from the primary database and registered in a secondary database. 4. A feature data created from an image image of a specific part of the registrant is registered in a database, and the feature data created from the image image of the person to be authenticated is registered in the database. In a personal authentication system for performing personal authentication by collating feature data, a primary database and a secondary database are provided in the database, all feature data are registered in the primary database, and a plurality of feature data of the same registrant are stored in the database. In addition to being able to register in the database, the number of matching matches and the number of matching matches are set for each feature data of all registrants. The feature data with the highest matching rate of The number of matches is updated, and the feature data other than the matching data is used as non-matching data.The number of matching non-matches of the feature data is updated, and the matching frequency is judged to be low by referring to the number of matching matches and the number of matching mismatches. A personal authentication system, wherein characteristic data is deleted from the primary database and registered in a secondary database. 5. A feature data created from an image image of a specific part of the registrant is registered in a database, and the feature data created from the image image of the person to be authenticated is registered in the database. In a personal authentication program for performing personal authentication by collating characteristic data, a procedure for registering a plurality of characteristic data of the same registrant in the database in a computer, and setting the number of times of matching and matching for each characteristic data of all registrants are set. A step of updating the number of times of matching and matching of the feature data with the feature data having the highest matching rate among feature data having a matching rate equal to or higher than a reference value as matching matching data in the matching of feature data at the time of personal authentication; Deleting from the database the characteristic data determined to be low in matching frequency from the number of matching matches. Personal authentication program, characterized in that. 6. A feature data created from an image image of a specific part of the registrant is registered in a database, and the feature data created from the image image of the person to be authenticated is registered in the database. In a personal authentication program for performing personal authentication by collating feature data, a procedure for registering a plurality of feature data of the same registrant in the database in a computer; In the step of setting the number of times, and the matching of the feature data at the time of personal authentication, the feature data having the highest matching rate among the feature data having the matching rate equal to or higher than the reference value is regarded as the matching matching data, and the matching matching number of the feature data is determined In addition to updating, the feature data other than the matching data is regarded as the matching nonconforming data, Personal authentication program, wherein the instructions new to, the procedure to remove the feature data from the database adapted frequency by referring to the matching incompatibility number and matching fit count is determined to be low, that is execution. 7. A feature data created from an image image of a specific part of a registrant is registered in a database, and the feature data created from the image image of a person to be authenticated is registered in the database. In a personal authentication program for performing personal authentication by collating feature data, a procedure for registering all feature data in a primary database in a computer, and enabling a plurality of feature data of the same registrant to be registered in the primary database, The procedure for setting the number of matching matches for every registrant's feature data, and the matching of feature data at the time of personal authentication. The procedure to update the number of matching matches of the feature data, and the matching frequency from the number of matching matches Personal authentication program for causing to execute a procedure for registering the secondary database feature data deleted from the primary database being determined to be low. 8. A feature data created from an image image of a specific part of the registrant is registered in a database, and the feature data created from the image image of the person to be authenticated is registered in the database. In a personal authentication program that performs personal authentication by comparing feature data, a procedure for registering all feature data in a primary database in a computer, and enabling a plurality of feature data of the same registrant to be registered in the database, Setting the number of matching matches and the number of matching matches for each registrant's feature data, and the feature data with the highest matching rate among feature data with matching rates equal to or higher than the reference value in the feature data matching at the time of personal authentication Is used as the matching data, the number of matching matches for the feature data is updated, and features other than the matching data are updated. The procedure for updating the number of times of non-conformity of matching of the feature data with the data as non-conforming non-conforming data; A personal authentication program characterized by causing a user to execute a procedure of registering in a secondary database.