JP2006228110A - Biometric authentication device, biometric authentication method, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a fingerprint authentication device or the like suitable for carrying out authentication as properly as possible even when extraction accuracy and collation accuracy of biometric characteristic points are low. <P>SOLUTION: In the fingerprint authentication device 101, a storing part 102 stores an identification code in coordination with a history of position information of characteristic points of a fingerprint extracted in the past, an input receiving part 103 receives inputs of an identification code to be authenticated, and a fingerprint image, an extracting part 104 extracts position information of characteristic points from the inputted fingerprint image, a collating pat 105 carries out collation of the position information of the extracted characteristic point by using each history of the position information corresponding to the inputted identification code as a collation object, an authenticating part 107 outputs that authentication of the identification code has succeeded when there is successful collation, and an updating part 108 updates the storing part 102 when authentication is successful, and it adds the extracted position information to the history of position information coordinated with the identification code. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a biometric authentication apparatus, a biometric authentication method, and a program for realizing these on a computer, which are suitable for performing authentication as appropriately as possible even when the extraction accuracy and verification accuracy of biometric feature points are low.

Conventionally, various techniques for authenticating a user have been proposed. The most widely used technique uses a recitation code such as a password or passphrase, and determines whether or not authentication is possible depending on whether or not a password or passphrase previously associated with the user has been input. Applications of such authentication techniques are disclosed in the following documents.
Japanese Patent No. 3324735

On the other hand, a fingerprint authentication technique using a fingerprint having a different pattern for each person has also been proposed. In such a technique, feature points such as end points and branch points of a raised portion (ridge) of a fingerprint are detected from a fingerprint image, and position information of the feature points is extracted, and a user's fingerprint registered in advance is also detected. The feature point location information and the extracted location information are compared to calculate the degree of similarity between the two feature points. Legal (sometimes called “Maneusha”) is also the beginning of usage.
In addition, a fingerprint recognition module and a fingerprint recognition device have been put on the market that accept input of a fingerprint image and collectively extract and output the position information of the feature points.
Furthermore, as a biometric recognition module or biometric recognition device other than fingerprints, information such as face and voice print, pupil iris, palm print, finger length and position, limb length and position, and blood vessel shape as information such as images and sounds There are also proposals for obtaining and using the feature points.

However, especially in the case of an inexpensive biological recognition module or biological recognition device, the extraction of positional information of biological feature points is not necessarily highly accurate. Further, depending on the pattern of the living body, the number of feature points may be small, and the accuracy of matching may be reduced. For this reason, there have been problems such as requiring the authorized user to input the living body many times or failing to authenticate the authorized user.
The present invention was made to solve the above-described problems, and is a biometric authentication device suitable for performing authentication as appropriately as possible even when the extraction accuracy and matching accuracy of biometric feature points are low, It is an object of the present invention to provide a biometric authentication method and a program that realizes these on a computer.

In order to achieve the above object, the following invention is disclosed in accordance with the principle of the present invention.
The biometric authentication apparatus according to the first aspect of the present invention includes a storage unit, an input reception unit, an extraction unit, a verification unit, an authentication unit, and an update unit, and is configured as follows.

That is, the storage unit stores an identification code and a history of position information of biological feature points extracted in the past (hereinafter, position information included in the history is referred to as “history position information”) in association with each other. To do.
Typically, the storage unit is configured by a non-volatile readable / writable storage medium such as a hard disk, an identification code such as a user number or a user name assigned to each user, and a biometric that has been successfully authenticated in the past for the user. One or more feature point position information is stored in association with each other as a history.

On the other hand, the input reception unit receives input of an identification code to be authenticated and biometric information.
Typically, the identification code is input using various keyboards and numeric keys, as well as reading a magnetic card or an IC (Integrated Circuit) card with a card reader, or reading a one-dimensional or two-dimensional barcode with a scanner. This is done by recognizing the code. On the other hand, input of biometric information such as a fingerprint is performed using a scanner that reads a user's biometric, a CCD (Charge-Coupled Device) camera, or the like. In addition, if a voiceprint is used, a device corresponding to biological information such as a microphone is adopted.

On the other hand, the extraction unit extracts the position information of the feature points from the biological information that has been accepted.
As described above, when fingerprints are used as biometric information, the positions of the end points and branch points of ridges that are characteristic points of the biometric are extracted together with the types thereof by image processing technology. In the case of a voiceprint, it may be possible to extract feature points of a frequency distribution graph of the voiceprint. Since inexpensive biological recognition modules and biological recognition devices are already available on the market, it is common to use them.

Further, the collation unit uses the history position information stored in the storage unit in association with the input identification code as a collation target, and extracts the position information of the extracted biometric feature points (hereinafter “extracted position information”). Are matched with each of the verification targets.
In the present invention, each position information of biometric feature points successfully matched in the past is used as a matching target. That is, if ten pieces of history position information are stored in the storage unit, the extracted position information is collated with each of the ten pieces of history position information. As a matching method, in the case of a fingerprint, a method similar to the minutiae method used in the existing biometric extraction matching method can be used. In addition, various methods corresponding to types of biological information can be used.

When the verification between the extracted position information and one of the verification targets is successful, the authentication unit outputs that the authentication of the identification code is successful.
Accordingly, when the 10 pieces of history position information and the extracted position information are collated in order, if the collation succeeds in the middle, it is not necessary to perform the remaining collation. At that time, authentication is successful.

On the other hand, when the authentication is successful, the update unit updates the storage unit so as to add the extracted position information to the history of the position information of the biometric feature point associated with the identification code that has been successfully authenticated.
By this update, the position information of the feature point of the biometric that has been successfully authenticated this time is used as history position information in the subsequent biometric authentication.

  According to the present invention, the location information of biometric feature points that have been successfully authenticated in the past is stored as a history, and the authentication is performed using the history in the next authentication. Even if the accuracy of the verification is low, the possibility that the authentication for the authorized user will fail can be reduced, and the user can be properly authenticated.

  In the biometric authentication device of the present invention, the storage unit further stores position information of predetermined biometric feature points (hereinafter referred to as “basic position information”) in association with the identification code, and a matching unit. The basic position information stored in the storage unit in association with the identification code accepted for input can be further set as a collation target.

For example, in the minutiae method using a fingerprint, first, the position information of the feature points of the user's biological body is registered, and authentication is performed by collating with the registered position information. The position information of the feature points of the living body is stored as basic position information in the storage unit separately from the history position information, and both are managed individually.
Then, both history position information and basic position information are used as collation targets.

The history position information is changed by a process performed by the update unit, but the basic position information is typically not changed thereafter once stored in the storage unit.
Therefore, when a fingerprint is used, the basic position information corresponds to the registered position information of the feature points of the living body in the conventional minutia method. Since the basic position information is position information serving as a reference for the user, it is desirable that biometric information is acquired by a highly accurate device and the position information of feature points is extracted by a highly accurate algorithm. However, position information acquired by a relatively inexpensive and low-accuracy device similar to that used in the input reception unit or extraction unit may be used.

  According to the present invention, position information of a plurality of living bodies for a certain user can be classified and managed into two types of basic position information that is not changed and history position information that can be changed, and a reference living body is determined. Therefore, biometric authentication can be performed appropriately.

In addition, the biometric authentication device of the present invention further includes a determination unit and can be configured as follows.
That is, the input receiving unit further receives an input of a recitation code.
Rather than using only biometric authentication, it asks for input such as a recitation number. Since many people use cash cards, there is little resistance from the user to input the recitation number. Therefore, an existing technique such as a recitation number is used as a supplementary technique for biometric authentication.

On the other hand, the determination unit determines whether or not the recitation code that has received the input is associated with the identification code that has received the input.
The determination process can use a technique similar to authentication using a general recitation number, password, or passphrase. For example, each user's recitation code is stored in the storage unit, and determination is made based on whether or not it matches the input recitation code, or a result obtained by performing predetermined conversion on the user's recitation code is stored in the storage unit. It is a technique such as determining whether the result of performing the conversion matches the input recitation code.

Furthermore, the authentication unit indicates that the verification of the identification code is successful when the verification by the verification unit is successful and the determination unit determines that the recitation code is associated with the identification code. Output.
That is, in the present invention, when the authentication by the collation of the extracted position information and the history position information (and the basic position information) is successful and the authentication by the recitation code is successful, the authentication is successful as a whole. is there.

  According to the present invention, by using authentication by recitation code as supplementary information, information that should not be added to the history, such as position information of a feature point of another person's biological body, is added due to an error in biometric authentication. This can be prevented as much as possible, and biometric authentication can be performed appropriately.

Further, in the biometric authentication device of the present invention, the update unit includes the history position information so that the number of history position information stored in the storage unit in association with the identification code that has been successfully authenticated does not exceed a predetermined threshold. Can be configured to remove any from
For example, when the predetermined threshold value is 1, the position information (and basic position information) of the feature point of the living body when the authentication is last successful is collated with the current position information.
According to the present invention, since the amount of history position information stored in the storage unit can be limited, an area such as a hard disk used as the storage unit can be appropriately distributed and managed.

Further, in the biometric authentication device of the present invention, the update unit can be configured to delete the oldest time stored in the storage unit when deleting any one from the history position information.
The present invention is one of the methods for limiting the number of history position information. In general, considering the fact that human behavior often uses the same device continuously, and that the technology of biometric information acquisition and feature point extraction can be advanced, the oldest one is deleted in preference to the new one. Adopting location information makes sense.

Further, in the biometric authentication device of the present invention, when deleting any one from the history of the history position information, the update unit succeeded in collating with the extracted position information after being stored in the storage unit. It can be configured to delete from those with a small number of times.
The present invention is one of the methods for limiting the number of history position information. There are many individual differences in biometric authentication devices and biometric authentication modules, and the resulting errors often show the same tendency. Therefore, history position information with a large number of successful collations in the past is considered highly likely to be successfully collated in the future, so delete from the few successful times and adopt new position information as history. That makes sense.

Further, in the biometric authentication device of the present invention, when deleting any one of the history position information from the history, the update unit can be configured to delete the one that is not most similar to the extracted position information.
The present invention is one of the methods for limiting the number of history position information. There are many individual differences in biometric authentication devices and biometric authentication modules, and the resulting errors often show the same tendency. Therefore, it is considered that the history position information similar to the latest position information is likely to succeed in the future, so delete the one with the lowest degree of similarity and adopt the new position information as the history. Is meaningful.

  The biometric authentication method according to another aspect of the present invention includes an identification code and a history of location information of biometric feature points extracted in the past (hereinafter, location information included in the history is referred to as “history location information”). Are executed in a biometric authentication apparatus having a storage unit, an input receiving unit, an extracting unit, a collating unit, an authenticating unit, and an updating unit, which are stored in association with each other, and an input receiving process, an extracting process, a collating process, an authenticating process, and an updating process And is configured as follows.

That is, in the input reception process, the input reception unit receives input of an identification code to be authenticated and biometric information.
On the other hand, in the extraction step, the extraction unit extracts the position information of the feature points from the biological information that has been accepted.
Further, in the collation step, the collation unit uses the history position information associated with the identification code accepted as input and stored in the storage unit as a collation target, and the positional information of the extracted biometric feature points (hereinafter “extraction”). "Completed position information") is collated with each of the collation targets.

In the authentication step, the authentication unit outputs that the authentication of the identification code is successful when the extracted position information is successfully verified with any of the verification targets.
On the other hand, in the update process, when the update unit succeeds in the authentication, the storage unit adds the extracted position information to the history of the position information of the biometric feature point associated with the identification code that has been successfully authenticated. Update.

A program according to another aspect of the present invention is configured to cause a computer to function as the biometric authentication device or to cause the computer to execute the biometric authentication method.
The program of the present invention can be recorded on a computer-readable information storage medium such as a compact disk, flexible disk, hard disk, magneto-optical disk, digital video disk, magnetic tape, and semiconductor memory.
The above program can be distributed and sold via a computer communication network independently of the computer on which the program is executed. The information storage medium can be distributed and sold independently from the computer.

  According to the present invention, a biometric authentication device, a biometric authentication method, and a program that realizes these on a computer are suitable for performing authentication as appropriately as possible even when the extraction accuracy and matching accuracy of biometric feature points are low. Can be provided.

  Embodiments of the present invention will be described below. In the following, for ease of understanding, an embodiment in which the present invention is realized using a computer will be described. However, the embodiment described below is for explanation and limits the scope of the present invention. It is not a thing. Therefore, those skilled in the art can employ embodiments in which each or all of these elements are replaced with equivalent ones, and these embodiments are also included in the scope of the present invention.

In particular, in order to facilitate understanding, a minutia method is employed in which feature points of fingerprints widely used in biometric authentication technology are extracted and collated from their positional relationships.
However, even in the technology using fingerprints, other collation technologies such as an image matching method for directly comparing and collating a registered fingerprint image with an inputted fingerprint image, and a cross section obtained by slicing a fingerprint image are regarded as waveforms. Thus, it is possible to apply a frequency analysis method in which a waveform spectrum series is used as feature information and a maximum addition is obtained and collated. In this case, both the information on the fingerprint image itself and the information on the waveform shape of the cross section obtained by slicing the fingerprint image can be considered as a kind of “position information of feature points of biological information”.

In addition, as a biometric recognition module or biometric recognition device other than fingerprints, face shape, voice print, pupil iris, palm print, finger length and position, limb length and position, blood vessel shape, etc. as information such as images and sounds It can also be applied to various biometric techniques such as acquiring and using the feature points.
Therefore, an embodiment to which an arbitrary biometric matching technique is applied is also included in the scope of the present invention.

FIG. 1 is a schematic diagram showing a schematic configuration of a fingerprint authentication device according to an embodiment of the present invention. Hereinafter, a description will be given with reference to FIG.
A fingerprint authentication apparatus 101 according to the present embodiment includes a storage unit 102, an input reception unit 103, an extraction unit 104, a collation unit 105, a determination unit 106, an authentication unit 107, and an update unit 108, and is configured as follows.

First, the following information is classified and stored in the storage unit 102 for each user.
(A) The identification code of the user. This is equivalent to a so-called login ID (IDentifier), which consists of a user number consisting only of numbers, a login name consisting mainly of alphanumeric characters, and the user's name and nickname using characters in other languages such as kana and kanji. Etc. that can identify each user.

  (B) Position information at the time of successful authentication among the position information of the feature points of the fingerprint extracted in the past for the user. Hereinafter, this is referred to as history position information. An upper limit may be set for the number of pieces of history position information stored in the storage unit 102 for a certain user. When the upper limit of the number is reached, the oldest ones are deleted, the frequently used history position information (the frequency of successful matching) is preferentially left in the storage unit 102, and the oldest ones are deleted. , Etc. can also be adopted.

  (C) Basic position information of feature points of the user's fingerprint. Typically, when the user is registered for the first time as a contrast for fingerprint authentication, position information extracted from the fingerprint image becomes basic position information. The basic position information may be omitted depending on the embodiment, and the position information at the time of first registration may be stored as one of the history position information.

Assuming that the upper limit of history position information is one, the position information of the feature point of the fingerprint when authentication was last successful for the user, that is, the latest position information is stored in the storage unit 102 as history position information. It will be.
Further, when the basic position information is not used, the position information of the feature points extracted when registering for the first time may be registered in the storage unit 102 as history position information.
Typically, the storage unit 102 is configured by a nonvolatile read / write storage medium such as a hard disk.

  FIG. 2 is a flowchart showing a control flow of fingerprint authentication processing executed by the fingerprint authentication apparatus according to the present embodiment. Hereinafter, a description will be given with reference to FIG.

When the fingerprint authentication process is started, first, the input receiving unit 103 of the fingerprint authentication apparatus 101 prompts the user to input the following information and receives these inputs (step S201).
(A) An identification code to be authenticated. It is an identification code assigned to the user.
(B) A fingerprint image of the user.
(C) A recitation code to be used for authentication. It corresponds to a so-called recitation number, password, and passphrase, and is associated with the identification code of the user in advance. Note that the input of the recitation code may be omitted depending on the embodiment.

  The input of the identification code and the recitation code is typically performed by various keyboards, numeric keys, touch panels, and the like. In addition, information may be input by reading a magnetic card or an IC card with a card reader or recognizing a barcode after reading a one-dimensional or two-dimensional barcode with a scanner. Moreover, it is good also as inputting an identification code and a recitation code by utilizing the device called a dongle.

  On the other hand, a fingerprint image is typically input using a scanner that reads a user's fingerprint, a CCD (Charge-Coupled Device) camera, or the like. Since various modules and devices are provided on the market for inputting a fingerprint image, these may be used as appropriate.

Next, the extraction unit 104 of the fingerprint authentication device 101 extracts feature point position information from the fingerprint image that has been accepted (step S202).
FIG. 3 is an explanatory diagram showing the positions of feature points in the minutiae method. Hereinafter, a description will be given with reference to FIG.

  The human fingerprint image 301 has a ridge 302 that is a raised portion of the skin of a human finger, and the ridge draws various patterns. Is a feature point. Further, the center of the ridge vortex may be employed as the feature point, or the ridge direction vector may be employed as the feature point position information.

  In the present embodiment, the positions of the end points and branch points of the ridges that are the characteristic points of the fingerprint are extracted together with their types by image processing technology. Since fingerprint recognition modules and fingerprint recognition devices that employ such a minutiae method have already been provided at a low price in the market, it is possible to extract feature point position information from a fingerprint image. The position information of feature points extracted in this way is referred to as “extracted position information”.

  When the fingerprint authentication apparatus 101 is configured using the above-described module or device, the module or device functions as the extraction unit 104. Since the feature point extraction corresponds to so-called image processing, when the fingerprint authentication apparatus 101 is configured using a computer, a CPU (Central Processing Unit) or an image processing processor of the computer is used. Functions as the extraction unit 104.

  Next, the collation unit 105 of the fingerprint authentication device 101 acquires the history position information (and basic position information) stored in the storage unit 102 in association with the identification code input in step S201 as a collation reference (step). S203).

  In conventional fingerprint authentication using the minutiae method, basic position information and extracted position information are compared, and it is determined whether or not the same type of feature point is at the same position from the position information of the feature point. Is going. That is, in the conventional minutiae method, only one fixed collation reference is collated with the extracted position information.

  On the other hand, in the fingerprint authentication according to the present embodiment, position information of feature points of fingerprints that have been successfully authenticated in the past is used as a collation reference. Therefore, the number of collation controls is not necessarily one, and can be plural. Further, the collation control may change as appropriate according to the history of authentication so far, and is not fixed.

  Subsequently, the collation part 105 repeats the following processes in order about the acquired collation control. That is, it is determined whether or not collation has been completed for all collation controls (step S204). If not completed (step S204; No), one collation control that has not been collated is selected (step S205). The collation reference and the extracted position information are collated (step S206). And it is judged whether collation was successful (step S207), and when it failed (step S207; No), it returns to step S204.

  As the matching technique performed in step S206, various minutiae matching techniques for matching by comparing the positional relationship of feature points can be used. Some fingerprint authentication modules and devices already available on the market provide a part for such verification processing separately as a program so that a new matching algorithm can be applied as appropriate. . In addition, in order to enable high-speed matching, there are some that constitute a dedicated electronic circuit. When these collation techniques are used, the CPU of the computer and the dedicated electronic circuit function as the collation unit 105 in cooperation with a hard disk or the like.

When the collation between the extracted position information and any collation reference is successful (step S207; Yes), the determination unit 106 of the fingerprint authentication device 101 associates the recitation code input in step S201 with the identification code. It is determined whether or not it has been received (step S208).
The determination process can use a technique similar to authentication using a general recitation number, password, or passphrase. For example, each user's recitation code is also stored in the storage unit 102, and it is determined whether or not it matches the input recitation code, or the result of performing a predetermined conversion on the user's recitation code is stored in the storage unit 102. Is stored, and a determination is made based on whether or not the result of applying the conversion to the input recitation code matches. Alternatively, a one-time password (abandoned password) may be employed. In addition, as in the credit card authentication, the determination unit 106 may make an inquiry to another authentication device.

  Therefore, typically, the CPU of the computer constituting the fingerprint authentication apparatus 101 functions as the determination unit 106.

When it is determined that the recitation code is associated with the identification code (step S208; Yes), the process repeats and the update unit 108 of the fingerprint authentication apparatus 101 uses the extracted location information as the user's information. The storage unit 102 is updated so as to be added as history position information associated with the identification code (step S209).
The location information of the feature points of the fingerprint that has been successfully authenticated this time will be used as history location information in subsequent fingerprint authentication, so the accuracy of extracting the feature points of the fingerprint is low or the accuracy of matching is low Even in this case, it is possible to reduce the possibility that the authentication with respect to the authorized user will fail, and the user can be properly authenticated.

  As described above, an upper limit may be provided for the number of pieces of history position information for a certain user. In this case, the update unit 108 deletes history position information that is less necessary. For example, the history position information stored in the storage unit 102 the oldest for the user may be deleted. As the result of collating the past extracted position information with each history position information, the number of successful times is stored in the storage unit 102, and the number of times that the number of times is successful (from the oldest when the number is the same) is deleted.

Humans often use the same device continuously, and the technology for acquiring fingerprint images and extracting feature points advances, and there are many individual differences in fingerprint authentication devices and fingerprint authentication modules, resulting in errors. Often change over time with the same trend. Accordingly, history position information that has been successfully collated in the past is considered highly likely to be successfully collated in the future. Therefore, in this embodiment, the number of successful registrations is deleted from the oldest registered one. The new location information is adopted as the history by deleting the ones with few.
Therefore, typically, the CPU of the computer constituting the fingerprint authentication apparatus 101 functions as the update unit 108 in cooperation with the hard disk or the like.

Then, the authentication unit 107 of the fingerprint authentication apparatus 101 outputs that the identification code has been successfully authenticated (step S210). This output result is used by another application program or the like, and the user can receive various services by being authenticated.
As described above, in this embodiment, it is assumed that the authentication of the identification code is successful when both the fingerprint verification and the verification by the recitation code succeed, but the determination unit 106 is appropriately omitted and the verification by the recitation code ( Step S208) may not be performed. However, if authentication based on recitation code is used supplementarily, it is possible to prevent as much as possible that information that should not be added to the history, such as position information of other person's fingerprint feature points, is added due to errors in fingerprint authentication. And fingerprint authentication can be performed appropriately.

  On the other hand, when it is determined that the recitation code is not associated with the identification code (step S208; No), the process returns to step S204.

Further, when collation is completed for all collation references (step S204; Yes), the authentication unit 107 of the fingerprint authentication apparatus 101 outputs that the authentication of the identification code has failed (step S211), and performs this processing. finish. This output result is used by another application program or the like, and the user cannot receive a desired service because the authentication has failed.
Therefore, typically, the CPU of the computer constituting the fingerprint authentication apparatus 101 functions as the authentication unit 107.

  As described above, in the present embodiment, the positional information of the feature points of the fingerprints that have been successfully authenticated in the past is stored as a history, and the verification is performed using the history in the subsequent authentication. Even when the accuracy of verification is low or the accuracy of verification is low, it is possible to reduce the possibility that authentication for a legitimate user will fail, and to perform user authentication appropriately.

  In the above-described embodiment, when there is something that succeeds in collation in the middle of sequentially comparing the history position information with respect to the extracted position information, the repetition is ended there, but collation may be performed to the end.

  In general, in matching two pieces of position information, a scale called similarity is obtained, and whether or not the matching is successful is determined based on the magnitude of this and a predetermined threshold.

  Therefore, when deleting any one of the history position information, if you want to delete in priority from the least number of successful matching in the past, or if you want to delete from the least similar to the extracted position information Therefore, it is necessary to perform a brute force comparison with all of the history position information.

  In order to change the repetition process in this way, the flag area prepared in the RAM or the like is cleared before the repetition is started (step S204), and the verification is successful (step S207; Yes), the fact is After recording in the flag area, the process returns to step S204.

  Then, after all collation (step S204; Yes), if it is stored in the flag area that the collation is successful, the process proceeds to step S208, and if not, the process proceeds to step S211.

  Various figure matching techniques can be used to calculate the similarity, but the simplest is to use the average of the distances between the corresponding feature points as the scale. To do. Then, it may be determined that the scale having the largest scale is deleted with priority.

  In addition, such biological information may form a cluster corresponding to the environment used. Therefore, with regard to history position information, clustering processing is performed on the basis of similarity, and the history position information that is most similar to the extracted position information among the history position information included in the cluster to which the extracted position information belongs is preferentially deleted. It is also good to do.

  In this way, even when a low-accuracy biometric authentication module or biometric authentication device is used, a pattern that is likely to be successfully authenticated is gradually accumulated in the biometric authentication device, and learning progresses and becomes more accurate. Biometric authentication can be performed.

  When such learning is performed, the input of the recitation code can be omitted as appropriate. That is, in step S201, the input of the recitation code is not required for the time being, and before proceeding to step S208, the degree of matching between the two is checked from the collation result between the current extracted position information and the history position information.

  When the average of the similarity between each history position information and the extracted position information indicates that the average is similar to a predetermined level or more, the variance of the similarity between each history position information and the extracted position information is predetermined. If it is within the range, the process proceeds to step S209 on the assumption that biometric authentication has been performed to a reliable level and that no input of the recitation code is required. Otherwise, the input of the recitation code is obtained, and if they match, the process proceeds to step S209, and if not, the process proceeds to step S211.

  According to the present embodiment, when there is little variation in biometric recognition and accurate recognition is possible, it is not necessary to ask the user to input an extra recitation code, while there are errors in various sensors, etc. Even if the learning is not sufficiently advanced, authentication is appropriately performed by inputting the recitation code so that the learning can be advanced, and the convenience of the user can be achieved.

  As described above, according to the present invention, a fingerprint authentication apparatus, a fingerprint authentication method, and a computer suitable for performing authentication as appropriately as possible even when the extraction accuracy and verification accuracy of fingerprint feature points are low Can be provided, and can be applied to various authentication technologies.

It is a schematic diagram which shows schematic structure of the fingerprint authentication apparatus which concerns on embodiment of this invention. It is a flowchart which shows the flow of control of the fingerprint authentication process performed with the fingerprint authentication apparatus which concerns on this embodiment. It is explanatory drawing which shows the position of the feature point in the minutiae method.

Explanation of symbols

DESCRIPTION OF SYMBOLS 101 Fingerprint authentication apparatus 102 Memory | storage part 103 Input reception part 104 Extraction part 105 Collation part 106 Judgment part 107 Authentication part 108 Update part 301 Fingerprint image 302 Ridge line 303 End point 304 Branch point

Claims (9)

A storage unit that stores an identification code and a history of position information of biometric feature points extracted in the past (hereinafter, position information included in the history is referred to as “history position information”) in association with each other;
An input receiving unit that receives input of an identification code to be authenticated and biometric information;
An extraction unit that extracts position information of feature points from the biometric information accepted as input;
The position information of the extracted feature points of the living body (hereinafter referred to as “extracted position information”) with the history position information stored in the storage unit associated with the input identification code received as a collation target. A matching unit that matches each of the matching targets,
An authentication unit that outputs that the authentication of the identification code is successful when the verification of the extracted position information and any of the verification targets is successful;
An update unit that updates the storage unit so as to add the extracted location information to the history of the location information of biometric feature points associated with the identification code that has been successfully authenticated when the authentication is successful. A biometric authentication device. The biometric authentication device according to claim 1,
The storage unit further stores position information (hereinafter referred to as “basic position information”) of predetermined biological feature points in association with the identification code,
The said collation part makes the basic position information matched with the identification code | symbol which received the said input memorize | stored in the said memory | storage part further as a collation object. The biometric authentication device according to claim 1 or 2,
The input receiving unit further receives an input of a recitation code,
A determination unit that determines whether or not the recitation code that has received the input is associated with the identification code that has received the input;
If the verification unit determines that the verification by the verification unit is successful, and the determination unit determines that the recitation code is associated with the identification code, the verification of the identification code is successful. That is characterized by the output of The biometric authentication device according to any one of claims 1 to 3,
The update unit deletes any one of the history position information from the history position information so that the number of history position information stored in the storage unit in association with the identification code that has been successfully authenticated does not exceed a predetermined threshold. The thing characterized by. The biometric authentication device according to claim 4,
When the update unit deletes any one of the history position information, the update unit deletes the oldest time stored in the storage unit. The biometric authentication device according to claim 4,
When deleting any one of the history position information from the history, the update unit deletes from the one that has been successfully collated with the extracted position information after being stored in the storage unit. A feature. The biometric authentication device according to claim 4,
When the update unit deletes any one of the history position information from the history, the update unit deletes the one that is most similar to the extracted position information. A biometric authentication method executed by a biometric authentication device having a storage unit, an input receiving unit, an extraction unit, a verification unit, an authentication unit, and an update unit,
In the storage unit, an identification code and a history of position information of biological feature points extracted in the past (hereinafter, position information included in the history is referred to as “history position information”) are stored in association with each other. And
An input receiving step in which the input receiving unit receives an input of an identification code to be authenticated and biometric information;
An extraction step in which the extraction unit extracts the position information of the feature points from the biometric information that has received the input;
The collation unit uses the history position information associated with the identification code that has received the input and stored in the storage unit as a collation target, and the positional information of the extracted biometric feature points (hereinafter “extracted position”). Information ”)) with each of the verification targets,
An authentication step of outputting that the authentication of the identification code is successful when the authentication unit succeeds in the verification of the extracted position information and any of the verification targets;
When the authentication is successful, the update unit updates the storage unit to add the extracted position information to a history of position information of biometric feature points associated with the identification code for which the authentication was successful. An updating process is provided. Computer
A storage unit that stores an identification code and a history of position information of biometric feature points extracted in the past (hereinafter, position information included in the history is referred to as “history position information”) in association with each other;
An input receiving unit that receives input of an identification code to be authenticated and biometric information;
An extraction unit that extracts position information of feature points from the biometric information accepted as input;
The position information of the extracted feature points of the living body (hereinafter referred to as “extracted position information”) with the history position information stored in the storage unit associated with the input identification code received as a collation target. A matching unit that matches each of the matching targets,
An authentication unit that outputs that the authentication of the identification code is successful when the verification of the extracted position information and any of the verification targets is successful;
When the authentication is successful, the update unit functions to update the storage unit so as to add the extracted location information to the history of the location information of biometric feature points associated with the identification code that has been successfully authenticated. A program characterized by letting

Images