JP4840036B2 - Biometric authentication apparatus and method - Google Patents

Description

  The present invention relates to a biometric authentication apparatus and method for performing personal authentication based on biometric information.

  In a biometric authentication system that performs personal authentication based on biometric information, feature point data (reference data) serving as a comparison reference is a server, a client PC (personal computer), a dedicated device, a sensor housing, a portable terminal such as an IC card, or the like. Stored in In general, the reference data of each individual is collectively stored in a predetermined storage area of such a device. Therefore, in the unlikely event that the reference data is leaked, there is a possibility that the biological information is stolen in its entirety.

  In addition, when a counterfeit / falsified device is used, there is a possibility of intrusion into a higher system by acting as if the device was authenticated even if the reference data was not stored in the device. For example, in response to an inquiry as to whether or not authentication is possible from a higher-level system, a device having a function of transmitting information “authentication OK” is manufactured and used even though authentication is not performed.

Further, as a technique related to the present invention, there is a technique called electronic tally (for example, Patent Document 1). In electronic tally, information is divided and stored and transmitted, and the information is restored by combining them when used. In this technique, the divided information can be restored only when all of the information is collected. Accordingly, by distributing the divided information, it is possible to make it difficult for information to be leaked or falsified. The electronic tally described in Patent Document 1 uses the data itself that has been divided and returned, but for example, when dividing biological information, the divided data is restored. It is not directly used at the time of verification of biometric authentication later, but is used as reference data. In biometric authentication, a comparison based on similarity is performed at the time of collation, not the perfect match of identity but how close it is.
JP 2000-845358 (WO 00/45358)

  The present invention has been made in consideration of the above circumstances, and prevents the leakage, theft, and misuse of feature point data (reference data), which is a reference for biometric authentication, and the prevention of impersonation due to tampering of the authentication device. It is an object of the present invention to provide a biometric authentication apparatus and method capable of achieving the above.

In order to solve the above-mentioned problem, the invention according to claim 1 reads a part of feature point data as a comparison reference and its error detection code from the first recording means, and the remainder is the first storage means. A second reading means provided in a device different from the first reading means, a part of the feature point data read from the first storage means, and the remainder of the feature point data read from the second storage means The point data is restored, and a combining means for confirming the presence / absence of an error using an error detection code, a generating means for generating feature point data from the captured biometric information, and a comparison reference combined without error by the combining means It is provided with the collation means which collates feature point data and the feature point data produced | generated by the production | generation means.

  The invention according to claim 2 is a portable storage unit in which the first storage unit stores a user-specific identification code together with the feature point data, wherein the second storage unit, the generation unit, and the collation The means is configured in the same server or is always connected, and the user-specific identification code stored in the first storage means is recorded in a predetermined storage means together with information indicating the result of the verification by the verification means. It further comprises recording means for performing.

  According to a third aspect of the present invention, the second storage unit, the generation unit, and the collation unit are configured or connected in the same computer, and no error is detected after combining by the combining unit. In addition, the generation of feature point data by the generation unit and the verification by the verification unit are performed.

  The invention according to claim 4 is characterized in that the feature point data serving as the comparison reference is formed from a plurality of types of biological information, and all the features related to at least one type of biological information are stored in the first storage means. Point data and error detection codes of feature point data stored in the second storage means are stored, and the second storage means stores all the feature point data relating to the remaining types of biological information and the An error detection code of feature point data stored in the first storage means is stored, and the combination means reads the feature point data read from the first storage means and the second storage means A plurality of types of feature point data are restored from the feature point data, and an error detection code is used to check whether or not each type of feature point data has an error. Feature point data is generated, and the collation unit collates, for each type, the feature point data that is the comparison reference combined without error by the combination unit and the feature point data generated by the generation unit. It is characterized by that.

  According to a fifth aspect of the present invention, the first storage means and the second storage means are both portable storage means, and the generation means and the verification means are configured in the same server or are always connected. It is characterized by being.

According to a sixth aspect of the present invention, a part of feature point data serving as a comparison reference and its error detection code are read from the first recording means, and the remaining part is provided in a device different from the first storage means. a reading step of reading from the second memory means has, along with restoring the feature data from the read and the remainder of the feature data from the first part and the second storage means of the feature data read from the storage means, Combining process for confirming the presence or absence of error using error detection code, generating process for generating feature point data from captured biometric information, feature point data as a reference for comparison without error in the combining process, and generating process And a collation process for collating the feature point data generated in (1).

  According to the first or sixth aspect of the present invention, the feature point data (reference data) serving as a comparison reference is divided and stored in the first storage unit and the second storage unit, and the first storage unit is stored by the combining unit. The feature point data is restored from a part of the feature point data read from the means and the remainder of the feature point data read from the second storage means, and the presence or absence of an error is confirmed using an error detection code. Reference data cannot be obtained unless there is a pair of feature point data stored in the first storage means and feature point data stored in the second storage means. Therefore, it is possible to prevent leakage, theft, and misuse of reference data, and to prevent impersonation due to tampering of the authentication device. Further, by dividing the reference data, the same effect as a tally can be obtained, and the validity of the person and the validity of the device to be used can be established by combining the divided data. That is, mutual authentication between the principal device and the system device can be easily established. Further, since the error detection code of the reference data is created before division and checked again at the time of combination, the validity of the divided data can be easily ensured.

  According to the second aspect of the present invention, for example, the first storage unit is configured as a portable storage device such as an IC card storing a user-specific identification code, and the second storage unit is configured to acquire biological information. It is provided on the server side that performs the verification process, and the authentication history can be registered and managed for each user in the recording means configured as a database in the server, so it can be used for entrance / exit / attendance management system and identity verification system. A suitable configuration can be easily constructed.

  According to the invention described in claim 3, since the second storage means, the generation means, and the collation means are configured or connected in the same computer, for example, the first storage means is configured as an IC card or a portable memory. The second storage means is configured as a storage device built in the personal computer, and the coupling means and the collating means are configured using a program, so that a configuration suitable for a personal computer logon system can be easily constructed. be able to.

  According to the invention described in claim 4, the feature point data serving as a comparison reference is formed from a plurality of types of biometric information, and all the features related to at least one type of biometric information are stored in the first storage means. The point data and the error detection code of the feature point data stored in the second storage means are stored, and the second storage means stores all the feature point data related to the remaining types of biological information and the first The error detection code of the feature point data stored in the storage means is stored, and the feature point data serving as a comparison reference is combined using the first and second storage means to perform authentication processing. Therefore, for example, a configuration suitable for a multimodal biometric authentication (multiple biometrics) system that combines face authentication and fingerprint authentication can be easily constructed.

  In the invention according to claim 5, since both the first storage means and the second storage means are portable storage means, and the generation means and the collation means are configured or always connected in the same server, For example, when the user carries the first storage means and the administrator of the system carries the second storage means, it is possible to easily construct a configuration suitable for the attending rented safe system.

  Embodiments of a biometric authentication system (biometric authentication apparatus) according to the present invention will be described below with reference to the drawings. The biometric authentication according to the present embodiment is roughly divided into two schemes: biometric information registration and authentication. First, a registration scheme will be described with reference to FIGS. 1 to 3, and then an authentication scheme will be described with reference to FIGS. 4 to 6.

  FIG. 1 is a system diagram showing a flow of information when registering biometric information according to the present embodiment. The terminal 1 is configured as a server, a client PC, a dedicated device for biometric authentication, a sensor housing for biometric information, and the like. It is configured with other storage devices. The storage medium A (2) is a portable terminal such as a portable memory including an IC card, a cellular phone terminal, a nonvolatile memory, and the like. The storage medium B (3) is a portable terminal similar to the storage medium A2, or a storage device built in another information processing apparatus such as the terminal 1.

  Processing at the time of biometric information registration according to the present embodiment will be described with reference to the system diagram of FIG. 1 and the flowcharts of FIGS. First, the terminal 1 scans biometric information such as fingerprints, palm prints, voice prints, irises, handwriting, and face types to be registered and loads them into a predetermined storage device of the terminal 1 (step S10 in FIG. 2). Next, feature point data 11 representing the biological information according to a predetermined condition is generated based on the captured biological information (step S11). The feature point data 11 is used as feature point data (reference data) that serves as a comparison reference in biometric authentication. Next, the feature point data 11 is divided (step S12).

  In the division of the feature point data 11 (step S12 in FIG. 2), first, as shown in FIG. 3, the length of the feature point data 11 is acquired (step S20). For example, it is required how many bytes the data length is. Next, a check code (error detection code) of the feature point data 11 is calculated. As a check code, in addition to a CRC (Cyclic Redundancy Check) code used in the present embodiment, a checksum, a hash function calculation result, or the like can be used. The check code (feature point data CRC14) obtained as a result of this calculation is stored in a predetermined storage device in the terminal 1.

  Next, the feature point data 11 is divided into feature point data A (12) and feature point data B (13) based on the length value obtained in step S20. In the present embodiment, the data is divided in half, and the feature point data A (12) is half (part) of the feature point data 11, and the feature point data B (13) is the remaining half (remaining) data. It becomes. For example, when the data length of the feature point data 11 is 256 bytes, the feature point data A (12) is converted into 128-byte data composed of odd-numbered bytes of data of 256 bytes. B (13) is divided into 128-byte data composed of even-numbered bytes of data. When the data length is an odd number, either data is 1 byte longer.

  Next, the divided data 12 and 13 and the check code 14 are stored in each target medium 2 and 3 in step S13 of FIG. In this example, feature point data A12 and feature point data CRC14 are stored in the storage medium A2, and feature point data B13 is stored in the storage medium B3.

  Next, processing at the time of authentication according to the present embodiment will be described with reference to the system diagram of FIG. 4 and the flowcharts of FIGS. First, the terminal 1 scans biometric information such as a fingerprint, a palm print, a voice print, an iris, a handwriting, and a face type to be authenticated and loads them into a predetermined storage device of the terminal 1 (step S30 in FIG. 5). Next, feature point data is generated based on the captured biological information (step S31). This feature point data is data to be compared with reference data in biometric authentication.

  Next, the divided data is read from the storage medium A2 and the storage medium B3 (step S32). In this example, feature point data A12 and feature point data CRC14 are read from storage medium A2, and feature point data B13 is read from storage medium B3 (FIG. 4).

  Next, the feature point data 11 is restored by combining the read divided data (step S33). In this combining process, first, each byte of the feature point data A12 is an odd number based on the total length of the feature point data A12 read from the storage medium A2 and the feature point data B13 read from the storage medium B3. The feature point data 11 is restored by combining them while rearranging them so that each byte of the feature point data B13 becomes an even-numbered byte (step S40 in FIG. 6). Next, as shown in FIG. 4, a check code 17 is calculated from the combined feature point data 11 and compared with the feature point data CRC 14 read from the storage medium A2, thereby confirming the CRC of the combined feature point data 11. To check whether there is an error (step S41).

  Next, in step S34 in FIG. 5, by comparing the feature point data obtained in step S31 by using the feature point data 11 obtained as a result of the combination without error in step S33 as reference data, between them, The similarity is determined, and when the similarity is equal to or greater than a predetermined value, it is determined that the collation has been successfully performed.

  According to the present embodiment, since the reference data is divided and cannot be used unless it is a pair, it is possible to prevent the reference data from being leaked, stolen, or misused. Moreover, by dividing the reference data, it is possible to obtain the effect of electronic tally, and to establish the validity of the principal and the validity of the device to be used by combining the divided data, so that mutual authentication can be easily established. . Further, since the check code of the feature point data of the biometric information is created before the division and checked again at the time of combination, the validity of the divided data can be ensured.

  Note that the reference data is divided and stored as (a) a self-device such as an IC card and another device such as an authentication terminal, and (b) a portable storage means such as an IC card. The device itself stores only the information indicating the storage location, and stores each divided data in two or more different storage locations of other different devices such as an authentication terminal, (c) an IC card or the like What is stored in another folder (another storage unit, another storage area in the same storage device) in its own device, (d) any of an electronic device having a server function, an electronic device having a client function, and a portable electronic device It is possible to store the data in two.

  Note that the divided feature point data is combined according to a predetermined method at the time of biometric information collation, but since the storage medium itself does not require a special configuration, the dividing method (combining method) is, for example, It can be freely set in advance on the system side according to the data length of the feature point data, the contents of the feature portion, and the like.

  Hereinafter, application examples of the embodiment described with reference to FIGS. 1 to 6 will be described with reference to FIGS. FIG. 7 shows an example of constructing an entrance / exit / attendance management and identity verification system using the biometric authentication device of the present invention. The entrance / exit / attendance management / identity confirmation system 100 is a system that performs identity verification by biometric authentication and manages entrance / exit and attendance / attendance. The storage medium 110 such as an IC card carried by each user stores a unique identification code together with the divided feature point data Bio_A 111 and CRC data Bio_CRC 112 of the feature data. The server 120 stores the information of the remaining remaining feature point data Bio_B121 in a predetermined storage device, and records information indicating each verification result of biometric authentication and date / time information corresponding to the identification code unique to the user. The database to be managed is connected and managed in the own apparatus or via a communication line. The server 120 executes biometric authentication and room entry / exit management by executing a predetermined program.

  At the time of verification of biometric authentication, first, the server 120 outputs a signal for reading out feature point data and the like from the storage medium 110 (step S101), and the storage medium 110 responds to the feature point data Bio_A111 and CRC data of the feature data. Bio_CRC 112, user-specific identification code information, and the like are output (step S102).

  The server 120 restores the feature point data (reference data) by combining the feature point data Bio_A 111 read from the storage medium 110 and the feature point data Bio_B 121 stored in its own device, and restores the feature point data (reference data) from the storage medium 110. Error detection using a check code is performed using CRC data Bio_CRC 112 of the read feature data, and it is determined whether or not the connection is normally performed (step S103).

  The server 120 performs biometric authentication after determining that the check code is normal, and compares the acquired biometric information feature point data with the restored feature point data, and if the predetermined similarity is obtained, the biometric authentication is successful. It is determined that the processing has ended (step S104). Next, the server 120 registers a user-specific identification code (ID) or the like in a predetermined beta base (recording unit), and ends the process at the time of collation.

  Next, an example of constructing a personal computer logon (PC logon) system will be described with reference to FIG. The PC logon system 200 is a system for performing identity verification by biometric authentication and enabling access to a predetermined system via the computer itself or a network. The storage medium 210 such as an IC card carried by each user stores a unique identification code together with the divided feature point data Bio_A 211 and CRC data Bio_CRC 212 of the feature data. The personal computer (PC) 220 stores the information of the remaining feature point data Bio_B 221 in a predetermined storage device. And PC220 performs the logon process by biometrics authentication by running a predetermined program.

  At the time of biometric verification, first, the PC 220 outputs a signal for reading feature point data and the like from the storage medium 210 (step S201). The information of the identification code unique to the user is output (step S202).

  The PC 220 restores the feature point data (reference data) by combining the feature point data Bio_A 211 read from the storage medium 210 and the feature point data Bio_B 221 stored in its own device, and reads it from the storage medium 210. Using the CRC data Bio_CRC 212 of the obtained feature data, error detection using a check code is performed to determine whether or not the connection has been normally performed (step S203).

  The PC 220 performs biometric authentication after determining the normality of the check code, compares the feature point data of the acquired biometric information with the restored feature point data, and if the predetermined similarity is obtained, the biometric authentication ends normally. It is determined that the log-on process has been performed, and a logon process is performed (step S204).

  Next, an example of constructing a multimodal biometric authentication system (multiple biometric authentication system) will be described with reference to FIG. The multimodal biometric authentication system 300 is a system that performs biometric authentication using a plurality of types of biometric information such as face authentication and fingerprint authentication.

  In this example, all (not divided) feature point data Bio_1 (311) relating to at least one type of biological information (for example, face information) and the storage means of the terminal 320 are stored in the storage medium 310 such as an IC card. In addition, an error detection code Bio_2_CRC 312 of feature point data Bio_2 (321) related to another type of biological information (for example, fingerprint information) is stored.

  Further, the terminal 320 configured using a computer or the like stores all feature point data Bio_2 (321) related to the remaining types of biological information (in this example, fingerprint information) in a predetermined storage area of the internal storage device. ) And the error detection code Bio_1_CRC 322 of the feature point data Bio_1 (311) stored in the storage medium 310.

  The terminal 320 performs biometric authentication using a plurality of types of biometric information by executing a predetermined program.

  At the time of verification of biometric authentication, first, the terminal 320 outputs a signal for reading feature point data and the like from the storage medium 310 (step S301), and the storage medium 310 responds to the feature point data Bio_1 (311) and feature data. Information such as CRC data Bio_2_CRC 312 is output (step S302).

  The terminal 320 restores the feature point data Bio_1 (311) read from the storage medium 310 and the feature point data Bio_2 (321) stored in the own device to the same directly accessible state as the original, and the storage medium 310 Error detection of the feature point data Bio_2 (321) using the CRC data Bio_2_CRC 312 of the feature data read out from the feature data, and error detection of the feature point data Bio_1 (311) using the CRC data Bio_1_CRC 322 of the feature data in the device itself To determine whether there is an error (step S303).

  The terminal 320 performs biometric authentication after determining that the check code is normal, collates the feature point data of the acquired multiple types of biometric information with the restored multiple types of feature point data for each type, and sets all types of biometric information. When a predetermined similarity is obtained, it is determined that biometric authentication has been normally completed (step S304).

  Next, an example of constructing a safe deposit system in a bank or the like will be described with reference to FIG. 10, the two storage media 410 and the storage medium 430 are both portable storage means such as an IC card, and the user carries the storage medium 410, and the storage medium 430 is managed by the system. This is a safe deposit box system in which the safe deposit box cannot be opened unless the two storage media 410 and 430 are used together. Such a safe deposit system is sometimes referred to as a bank attendance type safe deposit system.

  A storage medium 410 such as an IC card carried by each user stores a unique identification code together with the divided feature point data Bio_A 411 and CRC data Bio_CRC 412 of the feature data. In the storage medium 430 such as an IC card carried by the administrator side, the remaining feature point data Bio_B 431 is stored. Then, the terminal 420 performs a biometric authentication process by executing a predetermined program.

  At the time of verification of biometric authentication, first, the terminal 420 outputs a signal for reading feature point data and the like from the storage medium 410 and the storage medium 430 (step S401), and each storage medium 410, 430 responds to the feature point data accordingly. Bio_A411, CRC data Bio_CRC412 of feature data, feature point data Bio_B431, user-specific identification code information, and the like are output (steps S402 and S403).

  The terminal 420 restores the feature point data (reference data) by combining the feature point data Bio_A 411 read from the storage medium 410 and the feature point data Bio_B 431 read from the storage medium 430, and reads it from the storage medium 410. Error detection using a check code is performed using the CRC data Bio_CRC 412 of the feature data, and it is determined whether or not the combination is normally performed (step S404).

  The terminal 420 performs biometric authentication after determining that the check code is normal, and compares the acquired feature point data of the biometric information with the restored feature point data, and when the predetermined similarity is obtained, the biometric authentication is normally performed. It is determined that the user has finished, and the safe deposit box for the user is opened (step S405).

  The embodiment of the present invention is not limited to the above. For example, the number of divisions of reference data and the number of types of multimodal are set to three or more, a plurality of error detection codes are used, or error detection is performed. It is possible to appropriately change the code storage position to be moved or distributed to another storage medium. Further, the embodiment of the present invention includes a computer and a program executed by the computer, but the program can be distributed via a computer-readable recording medium or a communication line. is there.

The system figure for demonstrating the process at the time of biometric information registration in embodiment of the biometrics authentication system by this invention. The flowchart for demonstrating the flow of the process shown in FIG. The flowchart for demonstrating the flow of the feature point data division | segmentation process (S12) of FIG. The system figure for demonstrating the process at the time of biometric information collation in embodiment of the biometrics authentication system by this invention. The flowchart for demonstrating the flow of the process shown in FIG. The flowchart for demonstrating the flow of the division | segmentation data coupling | bonding process (S33) of FIG. Explanatory drawing for demonstrating the application example of this invention. Explanatory drawing for demonstrating the other application example of this invention. Explanatory drawing for demonstrating the other application example of this invention. Explanatory drawing for demonstrating the other application example of this invention.

Explanation of symbols

1 terminal (reading means, combining means, generating means, collating means)
2 Storage medium A (first storage means)
3 Storage medium B (second storage means)
S31 generating means S32 reading means S33 combining means S34 collating means

Claims (6)

A second memory provided in a device different from the device provided with the first storage means while reading a part of the feature point data serving as a comparison reference and its error detection code from the first recording means. Reading means to read from the means;
The feature point data is restored from a part of the feature point data read from the first storage means and the remainder of the feature point data read from the second storage means, and the presence or absence of an error is confirmed using an error detection code. A coupling means;
Generating means for generating feature point data from the captured biological information;
A biometric authentication device, comprising: a comparison unit that compares feature point data that is combined without error by the combining unit and that serves as a comparison reference with the feature point data generated by the generation unit. The first storage means is a portable storage means storing a user-specific identification code together with the feature point data,
The second storage unit, the generation unit, and the collation unit are configured or always connected in the same server,
2. The recording apparatus according to claim 1, further comprising: a recording unit that records a user-specific identification code stored in the first storage unit together with information indicating a verification result by the verification unit in a predetermined storage unit. Biometric authentication device. The second storage means, the generation means, and the collation means are configured or connected in the same computer,
The biometric authentication device according to claim 1, wherein when no error is detected after combining by the combining unit, generation of feature point data by the generating unit and verification by the verification unit are performed. The feature point data as the comparison reference is formed from a plurality of types of biological information,
The first storage means stores all feature point data related to at least one type of biological information and error detection codes of feature point data stored in the second storage means,
In the second storage means, all feature point data related to the remaining types of biological information and error detection codes of feature point data stored in the first storage means are stored,
The combining unit restores a plurality of types of feature point data from the feature point data read from the first storage unit and the feature point data read from the second storage unit, and each error detection code is used to restore each type of feature point data. Check whether there is an error in the type of feature point data,
The generation unit generates a plurality of types of feature point data from the plurality of types of captured biological information, and the verification unit uses the feature point data serving as a comparison reference combined without error by the combination unit, and the generation unit. The biometric authentication device according to claim 1, wherein the generated feature point data is collated for each type. Both the first storage means and the second storage means are portable storage means,
The biometric authentication device according to claim 1, wherein the generation unit and the verification unit are configured in the same server or are always connected. A second memory provided in a device different from the device provided with the first storage means while reading a part of the feature point data serving as a comparison reference and its error detection code from the first recording means. Reading process read from the means,
The feature point data is restored from a part of the feature point data read from the first storage means and the remainder of the feature point data read from the second storage means, and the presence or absence of an error is confirmed using an error detection code. The binding process;
A generation process for generating feature point data from captured biometric information,
A biometric authentication method comprising: a feature point data as a comparison reference that is joined without error in the joining process; and a matching step that matches the feature point data generated in the creation step.

Images