JP4445730B2 - Fingerprint authentication system and fingerprint authentication method - Google Patents

Description

  The present invention relates to a fingerprint authentication system and a fingerprint authentication method, and more particularly to improvements thereof.

  Personal authentication using fingerprints is characterized by the fact that fingerprints are universal and lifelong, and it is not necessary to carry them like an ID card or store them like a personal identification number. Realization of high security is expected.

  However, since personal authentication using fingerprints identifies each individual based on the feature amount of the fingerprint, it is not possible to extract a feature amount that can identify the individual depending on the fingerprint feature extraction method, and the fingerprint cannot be registered at the time of registration. There is a case.

  Therefore, in the conventional fingerprint authentication system, the personal data creation method is changed based on the evaluation value representing the quality of the personal data (dictionary data) created based on the fingerprint image (biological data) input at the time of registration. By changing the collation method, it is possible to register a user who has a fingerprint image that would otherwise be rejected (see, for example, Patent Document 1).

There is also a fingerprint authentication system that determines a fingerprint characteristic based on an input fingerprint image and selects a matching method (fingerprint matching algorithm) according to the fingerprint characteristic (see, for example, Patent Document 2).
JP 2000-99720 A (especially summary) JP 2003-44857 A (particularly summary)

  All of the above-described conventional fingerprint authentication systems determine the verification method on the system side, and thus appear to the user as a system that employs a single verification method. The same applies to those who attempt to obtain authentication illegally using forged fingerprint images or illegally obtained personal data. In other words, despite the fact that conventional fingerprint authentication systems have multiple fingerprint verification methods (fingerprint verification programs), no countermeasures have been taken to prevent unauthorized authentication by taking advantage of them, and reliability has been There is a problem that it hardly changes.

  SUMMARY OF THE INVENTION An object of the present invention is to provide a fingerprint authentication system in which a fingerprint authentication system having a plurality of fingerprint collation methods is provided with anti-fraud authentication measures taking advantage of the fingerprint authentication system and improved in reliability.

According to the present invention, a registration mode and a collation mode are provided. In the registration mode, the feature is extracted from the first fingerprint image input to the fingerprint input unit, and the first feature data representing the extracted feature is obtained. Registered in the database unit, and in the collation mode, the feature is extracted from the second fingerprint image input to the fingerprint input unit, and the second feature data representing the extracted feature is registered in the database unit In the fingerprint authentication system for collating with the first feature data, a plurality of selectable feature extraction methods and collation methods are displayed on the display unit in the registration mode and the collation mode, respectively, to the user. prompting selection of feature extraction methods and verification methods, as well as to specify a feature extraction method and verification method by manual input of a user, the said time of registration mode, the designated Fingerprint reader, characterized in that said first feature data without the inclusion of information that indicates a symptom extraction method and the verification method be registered in the database unit is obtained.

According to the present invention, in the registration mode, the feature is extracted from the first fingerprint image input to the fingerprint input unit, the first feature data representing the extracted feature is registered in the database unit, and collation is performed. In the mode, a feature is extracted from the second fingerprint image input to the fingerprint input unit, and second feature data representing the extracted feature is collated with the first feature data registered in the database unit. In the fingerprint authentication method, in the registration mode, a plurality of selectable feature extraction methods and verification methods are displayed on the display unit, prompting the user to specify the feature extraction method and verification method for registration, and using them who manually input using the specified feature extraction method by extracting the features from the first fingerprint image to obtain a first characteristic data, the information indicating the specified feature extraction method and verification method The registered in the database unit without Maseru, when the verification mode, and displays a plurality of feature extraction methods and verification methods selectable on the display unit, specifies the feature extraction method and verification method for matching to the user And extracting the feature from the second fingerprint image using a feature extraction method designated by manual input by the user to obtain the second feature data, and using a designated matching method A fingerprint authentication method characterized in that the second feature data is collated with the first feature data registered in the database unit.

Further, according to the present invention, in the registration mode, the feature is extracted from the first fingerprint image input to the fingerprint input unit, the first feature data representing the extracted feature is registered in the database unit, and collation is performed. In the mode, a feature is extracted from the second fingerprint image input to the fingerprint input unit, and second feature data representing the extracted feature is collated with the first feature data registered in the database unit. In the fingerprint authentication method, in the registration mode, a feature is extracted from the first fingerprint image using a plurality of combinations of a feature extraction method and a collation method, and each degree of detection is obtained. presented to the user by displaying on the display unit prompts to specify one from among the plurality of combinations, the user's manual input using the specified combined by the obtained first FEATURES data was registered in the database without the inclusion of information indicating the designated combination, when in the verification mode, and displays a plurality of feature extraction methods and verification methods selectable on the display unit, the user to encourage specifying the feature extraction method and verification method to obtain the second feature data by extracting the features from the second fingerprint image using a feature extraction method specified by the manual input of the user, In addition, a fingerprint authentication method is obtained in which the second feature data is collated with the first feature data registered in the database unit using a designated collation method.

  According to the present invention, the user can specify the feature extraction method and the collation method at the time of fingerprint registration and collation with the fingerprint authentication system, so that the data necessary for the authentication is distributed and unauthorized authentication is prevented. Reliability can be improved.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

  FIG. 1 shows a configuration of a fingerprint authentication system according to an embodiment of the present invention. The illustrated system includes a fingerprint input unit 11, a designation unit 12, a display unit 13, a database unit 14, and a control unit 15 connected thereto.

  The fingerprint input unit 11 has a finger placement surface for inputting a fingerprint image. The fingerprint input unit 11 generates input fingerprint image data representing the fingerprint shape (fingerprint image) of the finger placed on the finger placement surface and outputs it to the control unit 15.

  The designation unit 12 is a numeric keypad, for example. Alternatively, the designation unit 12 may be a touch panel combined with the display unit 13. In any case, the designation unit 12 receives an input from the user and outputs a signal corresponding to the input to the control unit 15. The designation unit 12 mainly designates a feature extraction method used for extracting features from the fingerprint image, and a collation method used for collating the extracted features with personal data registered in the database unit 14. Used to specify.

  The display unit 13 is an LCD panel or the like, and displays guidance for the user under the control of the control unit 15. The display unit 13 may be, for example, “Place your finger on the fingerprint input unit 11” or “Select a feature extraction method from the following: 1: ..., 2: ...”. In addition, a display prompting the user to input is performed.

  The database unit 14 stores personal data generated by the control unit 15 as will be described later.

  The control unit 15 includes a feature extraction / collation unit 151 and an authentication unit 152. The control unit 15 controls the fingerprint input unit 11, the designation unit 12, the display unit 13, and the database unit 14 to execute registration of personal data and fingerprint authentication using the registered personal data.

  The feature extraction / collation unit 151 extracts a feature that can be used for identifying the fingerprint from the input fingerprint image, obtains a degree of detection indicating the feature amount, and generates personal data of the extracted feature. Therefore, a plurality of different methods (algorithms or programs) are prepared in advance in the feature extraction / collation unit 151 as methods for extracting features from the input fingerprint image. The feature extraction / collation unit 151 collates the extracted feature with the feature (personal data) registered in the database unit 14 and obtains a collation degree indicating the degree of matching. For this purpose, the feature extraction / collation unit 151 includes a plurality of different methods (algorithms or programs) as collation methods for collating the features extracted from the input fingerprint image with the features registered in the database unit 14. Are prepared in advance.

  As a fingerprint collation method, a representative method includes a minutia method and an image correlation method. There are various types of these methods. Any method can be used for the fingerprint authentication system according to the present embodiment.

  The feature extraction / collation unit 151 executes a method selected from a plurality of prepared feature extraction methods and a plurality of collation methods by a signal from the designation unit 12. In the following description, it is assumed that a plurality of feature extraction methods and a plurality of matching methods are preliminarily combined in a one-to-one manner, and the selection is performed in units of sets.

  When a signal indicating the illuminance level is sent from the feature extraction / collation unit 151, the authentication unit 152 compares the collation level indicated by the illuminance level with a preset threshold value and outputs the result as an authentication result.

  Hereinafter, the operation of the fingerprint authentication system of FIG. 1 will be described.

  First, the operation at the time of fingerprint registration will be described with reference to FIGS.

  A user who intends to perform fingerprint registration (hereinafter referred to as a registrant) performs a predetermined operation on the designation unit 12 to set the operation mode of the fingerprint authentication system to the fingerprint registration mode.

  Next, the registrant places a finger on the finger placement surface of the fingerprint input unit 11 and inputs a fingerprint image for registration (step S201). The fingerprint input unit 11 generates input fingerprint image data 301 representing the input fingerprint image and outputs it to the feature extraction / collation unit 151.

  Upon receiving the input fingerprint image data from the fingerprint input unit 11, the control unit 15 displays a plurality of selectable feature extraction / collation methods on the display unit 13 and selects a feature extraction / collation method for the user. Prompt. When the registrant does not input a number or the like for selecting any one or more of the plurality of feature extraction / collation methods presented within a predetermined time into the designation unit 12, or indicates non-selection When a number or the like is input to the designation unit 12 (when not selected in step S202), the control unit 15 sets all methods as the selected feature extraction / collation method (step S203).

  When the registrant inputs a number or the like for selecting one or more of a plurality of feature extraction / collation methods presented within a predetermined time into the specification unit 12 (when selecting in step S202) A signal 302 indicating a number selected and designated from the unit 12 is supplied to the control unit 15. The control unit 15 sets a feature extraction / collation method corresponding to the number indicated by the signal from the designation unit 12 as the feature extraction / collation method selected and designated (step S204).

  The feature extraction / collation unit 151 performs feature extraction processing on the fingerprint image indicated by the input fingerprint image data from the fingerprint input unit 11 by using the selected feature extraction / collation method (step S205).

  After executing all the methods set as the selected feature extraction / collation method, the feature extraction / collation unit 151 obtains the degree of detection corresponding to each method, and displays a signal 303 indicating the obtained degree of detection on the display unit 13. To supply. The display unit 13 displays the degree of detection of each method (step 206), and displays to prompt the selection and specification of the feature extraction / collation method at the time of registration.

  The registrant refers to the degree of detection of each method displayed on the display unit 13 and determines a feature extraction / collation method at the time of registration. Then, the registrant designates the feature extraction / collation method determined by the designation unit 12 (step S207). The designation unit 12 supplies a signal indicating the designated feature extraction / collation method to the control unit 15.

  Upon receiving a signal indicating the specified feature extraction / collation method from the designation unit 12, the control unit 15 generates personal data 304 including information indicating the characteristics of the fingerprint image extracted by the corresponding feature extraction / collation method. (Step S208). The personal data does not contain information indicating the feature extraction method used (not associated), and is processed such as encryption so that the feature extraction method cannot be specified. Shall. The generated personal data 304 is sent to and stored in the database unit 14 (step S209).

  As described above, personal data for fingerprint authentication (or personal authentication) is registered in the database unit 14.

  Next, an operation at the time of fingerprint authentication of the fingerprint authentication system of FIG. 1 will be described with reference to FIGS.

  First, a user who wants to receive authentication (hereinafter, a person to be authenticated) places a finger used for authentication on the finger placement surface of the fingerprint input unit 11 (step S401). The fingerprint input unit 11 generates input fingerprint image data 501 representing the fingerprint image of the finger placed on the finger placement surface, and outputs it to the control unit 15. Note that the operation mode of this fingerprint authentication system is normally the fingerprint authentication mode, so there is no need to specify it again.

  The feature extraction / collation unit 151 that has received the input fingerprint image data from the fingerprint input unit 11 displays a display that prompts the user to select one of a plurality of feature extraction / collation methods. 13 is displayed.

  When the person to be authenticated selects any feature extraction / collation method from the designation unit 12 (step S402), a signal 502 indicating the selected feature extraction / collation method is sent to the feature extraction / collation unit 151.

  The feature extraction / collation unit 151 that has received the signal 502 indicating the selected feature extraction / collation method performs feature extraction on the input fingerprint image indicated by the input fingerprint image data using the selected feature extraction / collation method. Do. Subsequently, the feature extraction / collation unit 151 reads the personal data 503 registered in the database unit 14 and collates with the features included therein (step S403). Then, as a result of the collation, the feature extraction / collation unit 151 outputs a collation degree 504 indicating the degree of coincidence of those features to the authentication unit 152.

  The authentication unit 152 compares the threshold corresponding to the feature extraction / collation method executed by the feature extraction / collation unit 151 with the collation degree to determine whether the person to be authenticated and the registrant are the same person. Is determined (step S404). And the authentication part 152 displays the authentication result on the display part 13 (step S405).

  As described above, personal authentication by fingerprint is performed.

According to the present embodiment, when the feature extraction / collation method designated by the non-authenticator at the time of fingerprint authentication is different from the feature extraction / collation method designated at the time of registration, the collation degree decreases. As a result, even if the person to be authenticated is the same person as the registrant, authentication is denied. That is, in this fingerprint authentication system, authentication cannot be obtained unless the feature extraction / collation method designated at the time of registration is stored. Therefore, authentication cannot be obtained only by forging a fingerprint image or stealing personal data, and unauthorized authentication is prevented.

  Although the present invention has been described with reference to the preferred embodiments, the present invention is not limited to the above embodiments.

  For example, in the above-described embodiment, manual input is performed from the specifying unit 12 in order to select and specify the feature extraction / collation method at the time of authentication, but information representing the feature extraction / collation method selected and specified at the time of registration is stored in the IC. It may be recorded on a recording medium such as a card, and the information may be read by a reading device such as an IC card reader at the time of authentication. In addition to or instead of the information indicating the feature extraction / collation method selected and specified at the time of registration, a program (part) or a parameter (part) necessary for executing the method is stored in an IC card or the like. You may make it record on a recording medium.

  Moreover, in the said embodiment, ID etc. are not utilized, However, You may combine with ID etc. Not only can the reliability be improved by the combination with the ID, but also the identification of personal data that is symmetric with respect to the collation can be facilitated and a higher-speed system can be obtained.

  Furthermore, in the above-described embodiment, the feature extraction method and the collation method are combined in advance. However, these may be specified independently or only one of them can be designated.

  Furthermore, in the above-described embodiment, various guidance tables are provided to the user using the display unit 13, but voice guidance using a speaker or the like may be performed together with or instead of the display unit. Good.

It is a block diagram which shows the structure of the fingerprint authentication system which concerns on one embodiment of this invention. 3 is a flowchart for explaining an operation in a registration mode of a fingerprint authentication unit in FIG. 1. It is a schematic diagram for demonstrating the flow of the signal in the registration mode of the fingerprint authentication part of FIG. 3 is a flowchart for explaining an operation in an authentication mode of a fingerprint authentication unit in FIG. 1. It is a schematic diagram for demonstrating the flow of the signal in the authentication mode of the fingerprint authentication part of FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 11 Fingerprint input part 12 Specification part 13 Display part 14 Database part 15 Control part 151 Feature extraction and collation part 152 Authentication part

Claims (7)

It has a registration mode and a verification mode.
When in the registration mode, the feature is extracted from the first fingerprint image input to the fingerprint input unit, and the first feature data representing the extracted feature is registered in the database unit,
In the collation mode, the feature is extracted from the second fingerprint image input to the fingerprint input unit, and the second feature data representing the extracted feature is registered with the first feature data registered in the database unit. In the fingerprint authentication system to collate,
In the registration mode and the collation mode, a plurality of selectable feature extraction methods and collation methods are displayed on the display unit, respectively, and the user is prompted to select the feature extraction method and collation method. The feature extraction method and the matching method are specified by manual input , and
The fingerprint authentication system , wherein in the registration mode, the first feature data is registered in the database unit without including information indicating the specified feature extraction method and the matching method . A designating unit for designating the feature extraction method and the matching method;
A feature extraction / collation unit that extracts features from the first or the second fingerprint image by the feature extraction method and collation method designated by the designation unit, and outputs a detection degree or a collation degree;
The fingerprint authentication system according to claim 1 , further comprising: The fingerprint authentication system according to claim 2 , wherein the feature extraction method and the collation method are selected and specified from a plurality of predetermined combinations. The fingerprint authentication system according to claim 2 , wherein the feature extraction method and the collation method are individually selected and specified. When in the registration mode, the feature is extracted from the first fingerprint image input to the fingerprint input unit, and the first feature data representing the extracted feature is registered in the database unit,
In the collation mode, a feature is extracted from the second fingerprint image input to the fingerprint input unit, and second feature data representing the extracted feature is collated with the first feature data registered in the database unit. In the fingerprint authentication method to
In the registration mode, the display unit displays a plurality of selectable feature extraction methods and collation methods, and prompts the user to specify the feature extraction method and collation method for registration.
Information indicating the specified feature extraction method and collation method by extracting the feature from the first fingerprint image using the feature extraction method specified by the user's manual input to obtain the first feature data Registered in the database part without including
In the collation mode, a plurality of selectable feature extraction methods and collation methods are displayed on the display unit to prompt the user to specify a feature extraction method and collation method for collation,
The feature is extracted from the second fingerprint image using a feature extraction method designated by manual input by a user to obtain the second feature data, and the second feature data is obtained using a designated matching method. A fingerprint authentication method, wherein feature data is collated with the first feature data registered in the database unit. When in the registration mode, the feature is extracted from the first fingerprint image input to the fingerprint input unit, and the first feature data representing the extracted feature is registered in the database unit,
In the collation mode, a feature is extracted from the second fingerprint image input to the fingerprint input unit, and second feature data representing the extracted feature is collated with the first feature data registered in the database unit. In the fingerprint authentication method to
In the registration mode, a feature is extracted from the first fingerprint image using a plurality of combinations of a feature extraction method and a matching method, and each degree of detection is obtained.
Display the obtained degree of detection on the display unit and present it to the user, prompting to specify one of the plurality of combinations,
The first feature data obtained by using the specified combinations by the user manually inputs, registered in the database without the inclusion of information indicating a combination of the specified,
In the collation mode, a plurality of selectable feature extraction methods and collation methods are displayed on the display unit, prompting the user to specify the feature extraction method and collation method,
The feature is extracted from the second fingerprint image using a feature extraction method designated by manual input by a user to obtain the second feature data, and the second feature data is obtained using a designated matching method. A fingerprint authentication method, wherein feature data is collated with the first feature data registered in the database unit. 7. The fingerprint authentication method according to claim 5, wherein the first characteristic data is encrypted and registered in the database.

Images