JP5402137B2 - Biometric authentication device - Google Patents

Description

  The present invention relates to a biometric authentication technique for performing personal authentication using human body features such as fingerprints and veins.

  In modern society, identity verification is required in various situations, and biometric authentication that uses some features of the human body, such as fingerprints, veins, facial images, irises, etc., as one method for accurately identifying the identity of people in recent years. It is spreading. For example, biometric authentication is used to determine whether to enter or leave a restricted area where only authorized persons can enter, whether to log in to a personal computer, whether to access various services such as online transactions, and the like.

  Such identification by biometric authentication is roughly divided into two types of usage. The first is to present an ID card or enter a personal ID, password, password, etc. before biometric authentication. This is a usage mode in which biometric authentication is performed after determining whether or not there is. In biometric authentication, it is determined whether the user is really the person or another person. In this usage mode, a person to be authenticated is determined in advance, and the biometric information previously registered by the target person and the biometric information newly presented at the time of authentication are associated one-to-one to perform biometric authentication. Generally, this is called one-to-one authentication.

  The other is a usage mode in which biometric authentication is performed immediately without presenting any information for confirming the user before performing biometric authentication, unlike the above usage mode. Since the user has not been determined in advance, the role of biometric authentication in this usage mode is whether the biometric information of the user is registered in the database for verification and if it is registered in the database for verification It is required to determine who the person is. Since there are usually a plurality of biometric information registered in the database for verification and the biometric information of one person to be authenticated is comprehensively verified against a plurality of biometric information registered in the database for verification, This form of use is called one-to-N authentication or one-to-many authentication.

JP 2005-227933 A

  The one-to-one authentication described above imposes a burden on the user, such as the presentation of an ID card and the input of a personal ID / password / password, etc. When implementing biometric authentication in situations used by many ordinary people In many cases, one-to-N authentication is required.

  Compared to the one-to-one authentication in which a biometric authentication such as a fingerprint is presented by first presenting a personal ID and the like, and then a biometric such as a fingerprint is presented, a biometric such as a fingerprint is presented from the beginning without establishing an individual beforehand One-to-N authentication that performs biometric authentication has a fundamental problem with this method. In other words, as the number of biometric information registered in the verification database increases, the time required for biometric authentication becomes longer.

  This is because one-to-N authentication basically requires one user to collate with all the biometric information of all users registered in the collation database. The degree to which this time increases depends on the biometric algorithm, and various speedups have been made by devising the biometric algorithm. However, if the number of biometric authentication registration data increases, the time required for verification will increase. The fact that it increases is unchanged.

  If the data to be verified can be narrowed down from the registered data, it can be expected to shorten the processing time of the 1-to-N authentication. For example, one-to-one authentication is an extreme example in which the personal ID is presented in advance to narrow down the data to be collated to one person. A method equivalent to this is to allow the user to present and input some information in advance. There is a method to narrow down the data to be verified to a certain number. However, these methods are to allow the user to perform a special operation, and are inconsistent with the selection of the 1-N authentication usage mode in order not to impose a burden on the user.

  On the other hand, Patent Document 1 discloses a technique for specifying which living body part from an image captured by a biosensor by a method such as image processing and performing a matching process on the specified living body part. However, since the living body part is uniquely determined, a heavy processing is required for the determination, and the time for the matching process becomes longer for recovery when the specified living body part is wrong. There are inconveniences such as.

  In view of the above-described conventional problems, the user is burdened with an extra burden while maintaining the use form of the one-to-N authentication in which the identity verification is performed only by the basic operation of the biometric authentication that presents a living body such as a fingerprint or a vein. It is an object of the present invention to provide a biometric authentication device that can narrow down the data to be verified and reduce the time for verification processing.

In one embodiment of this biometric authentication device, biometric information holding means for holding biometric information about a plurality of biometric parts of a plurality of authentication subjects as registered biometric information so that the authentication target person and the biometric part can be specified; Biometric information acquisition means for acquiring biometric information of a user as authentication biometric information, feature quantity acquisition means for acquiring a feature quantity indicating how to present the biometric part during authentication, and a predetermined biometric part based on the feature quantity A part estimation probability acquisition unit that acquires a part estimation probability indicating certain probability, a collation control unit that determines a resource allocation of a collation process for each biological part based on the part estimation probability and controls execution of the collation process; based on resource allocation and a collating means that performs collation processing of the registered biometric information and the authentication biometric information, said verification control means, Li corresponding to the acquired part estimation probability Assigned to each body part to over scan rate at the start of the verification process, the verification process for each body part by the time of collation processing ends is completed, determining a continuous resource allocation.

  In the disclosed biometric authentication device, by using effectively auxiliary information that has not been effectively used in the past, identity verification is performed only by a basic operation of biometric authentication in which a biometric such as a fingerprint or a vein is presented 1 While maintaining the use form of N authentication, the user is not burdened excessively, the data to be collated can be narrowed down, and the collation processing time can be shortened.

It is a figure which shows the structural example of the biometrics apparatus concerning one Embodiment. It is a figure which shows the example of the difference in the angle in the imaging | photography surface which is a feature-value at the time of targeting the palm of the right hand and the palm of the left hand. It is a figure which shows the example of the difference in the angle or distance with respect to the imaging | photography surface which is a feature-value at the time of making the palm and the back of a hand into object. It is a figure which shows the example of the difference in the angle in the imaging | photography surface which is a feature-value at the time of making the finger of the right hand and the finger of the left hand into object. It is a figure which shows the example of the difference in the pupil position which is a feature-value at the time of making right eye and left eye into object. It is a flowchart which shows the process example of embodiment. It is a figure which shows the example of biometric information template data. It is a figure which shows the example which acquires the angle in the imaging surface of a palm by image processing. It is a figure which shows the example which acquires the angle in the imaging | photography surface of a palm with the touch sensor as a feature-value sensor. It is a figure which shows the example of feature-value and site | part estimation probability corresponding | compatible data. It is a figure which shows the example of resource allocation. It is a flowchart which shows the example of the collation process which considered resource allocation.

  Hereinafter, preferred embodiments of the present invention will be described.

<Configuration>
FIG. 1 is a diagram illustrating a configuration example of a biometric authentication apparatus according to an embodiment.

  In FIG. 1, a biometric authentication device 1 includes a biometric sensor unit 2, a biometric information acquisition unit 3, a feature amount acquisition unit 4, a part estimation probability acquisition unit 5, a feature quantity / partial estimation probability correspondence data holding unit 6, and a priority matching control unit. 7, a biometric information template collation unit 8, and a biometric information template data holding unit 9.

  The biometric sensor unit 2 captures features of a human body such as fingerprints, veins, facial images, and irises to acquire image information, and a feature that directly acquires feature amounts from a part of the human body. A quantity sensor 22. It should be noted that the feature quantity sensor 22 is not necessary when the feature quantity is obtained by image processing from the image information obtained by the living body imaging sensor 21.

  Here, the feature amount is a value indicating how the user of the biometric authentication device 1 presents his / her own biological part to the biometric sensor unit 2. This is because when the user presents his / her biological part to the biological sensor unit 2, the biological part presented to the biological sensor unit 2 is naturally and inevitably presented due to movement restrictions due to the physical structure of the person. We are taking advantage of changes in ways. This feature amount is not information for each individual for specifying the individual, but information indicating the tendency of the entire user.

  Specifically, for example, if the target is the palm of the right hand and the palm of the left hand, the angle of the hand in the imaging plane, if the target is the palm and the back of the hand, the angle or distance of the hand relative to the imaging plane, the right and left fingers If there are, the angle of the finger in the imaging plane, the position of the pupil in the imaging image, etc. can be used for the right eye and the left eye.

  FIG. 2 is a diagram showing an example of the difference in angle within the imaging plane, which is a feature amount when the palm of the right hand and the palm of the left hand are targeted, and is applied to an automatic ticket gate targeting the vein pattern of the palm. It is shown as an example. FIG. 2A is a view of a plurality of automatic ticket gates as viewed from above, and FIGS. 2B and 2C are views of the automatic ticket gate as viewed from an oblique direction. In this example, it is assumed that the hand is presented to the right biological sensor unit 2 (2A, 2B) with respect to the traveling direction of the automatic ticket gate.

  Normally, it is common to register biometric information for both hands in case you have an injury to your hand or if you have a baggage and cannot use one hand, and this example also uses an automatic ticket gate. It is assumed that the user registers biometric information on the right hand and left hand in advance.

  2 (a) and 2 (b), if the user U1 presents the right hand to the biosensor unit 2A, the user U1 presents the hand on the same side as the biosensor unit 2A. It becomes a natural state. 2A and 2C, if the user U2 presents the left hand to the biosensor unit 2B, the user U2 presents the hand opposite to the biosensor unit 2B. And hold your hand a little sideways. Therefore, it is possible to use the angle of the hand in the imaging surface by the biosensor units 2A and 2B as a feature amount for estimating the right hand or the left hand.

  FIG. 3 is a diagram showing an example of the difference in angle or distance with respect to the imaging plane, which is a feature amount when the palm and the back of the hand are targeted, and the living body sensor unit 2 installed on the wall presents the palm or the back of the hand. This is an example of performing authentication. As shown in FIGS. 3 (a) and 3 (b), there is a difference between the angle and the distance with respect to the biosensor unit 2 due to the movement limitation of the wrist when the palm is presented and the back of the hand is presented. That is, as shown in FIG. 3A, when presenting the palm, the degree of freedom is relatively large, and it is possible to take a posture from a state where the hand is bent slightly backward to a state where the hand is bowed forward. However, when the back of the hand is presented as shown in FIG. 3 (b), the degree of freedom is relatively small and the posture of warping backward is difficult, and the posture of bowing slightly forward becomes natural. Therefore, it is possible to use the angle or distance of the hand with respect to the imaging surface by the biosensor unit 2 as a feature amount for estimating the palm or the back of the hand.

  FIG. 4 is a diagram illustrating an example of a difference in angle within the imaging plane, which is a feature amount when a right finger and a left finger are targeted. Physical restriction in the case of presenting the finger of the right hand to the biosensor unit 2 as shown in FIG. 4A and the case of presenting the finger of the left hand to the biosensor unit 2 as shown in FIG. Therefore, a difference occurs in the angle with respect to the living body sensor unit 2 naturally. Therefore, it is possible to use the angle of the finger within the imaging surface by the biosensor unit 2 as a feature amount for estimating the finger of the right hand or the finger of the left hand.

  FIG. 5 is a diagram illustrating an example of a difference in pupil position, which is a feature amount when the right eye and the left eye are targets. In the case of presenting the right eye as shown in FIG. 5 (a) and the case of presenting the left eye as shown in FIG. 5 (b), the relationship between the installation position of the biosensor unit (2) and the standing position of the user. Naturally, there is a difference in the position of the pupil in the captured image of the biosensor unit (2). Therefore, the position of the pupil in the captured image by the biosensor unit 2 can be used as a feature amount for estimating the right eye or the left eye.

  It should be noted that other information may be used as long as it satisfies the point that the way of presenting the living body part to be presented to the living body sensor unit 2 naturally and inevitably changes due to the movement restriction due to the human physical structure. For example, the distance between the biological sensor part 2 and the biological part when the biological part is presented to the biological sensor part 2, the pressure applied to the biological sensor part 2 when contacting the biological sensor part 2, and the biological part on the biological sensor part 2 This is the speed at the time of presentation. Taking the left and right hands as an example, the hand farther away from the biosensor unit 2 is held at a position that is stochastically separated from the hand closer to the biosensor unit 2 when presented to the biosensor unit 2. There are many. In the case of a configuration in contact with the biosensor unit 2, the pressure applied to the biosensor unit 2 is greater when a biopart closer to the biosensor unit 2 is used than when a biopart far from the biosensor unit 2 is used. growing. A biological part close to the biological sensor unit 2 is quickly held over, but when a remote biological part is held over, it becomes slow.

  Returning to FIG. 1, the biometric information acquisition unit 3 acquires biometric information (authentication biometric information) indicating biometric features such as vein patterns and fingerprint patterns from image information acquired by the biometric imaging sensor 21 of the biosensor unit 2. have. The biometric information is binary image data or coordinate value data of feature points (pattern vertices, branch points, end points, etc.).

  The feature amount acquisition unit 4 performs image processing on the image information acquired by the living body imaging sensor 21 of the biosensor unit 2 or processes the feature amount acquired by the feature amount sensor 22 as it is or as described above. It has the function to acquire.

  The part estimation probability acquisition unit 5 obtains the feature quantity / partial estimation probability correspondence data in which the feature quantity grasped in advance by a preliminary experiment or the like is associated with the part estimation probability from the feature quantity / partial estimation probability correspondence data holding unit 6. It has a function to acquire a site estimation probability by referring to as a key. The part estimation probability is a value indicating the certainty of which biological part is determined from the feature amount. An example of the feature amount / part estimation probability correspondence data holding unit 6 will be described later.

  The priority collation control unit 7 determines the resource allocation of the collation processing for each biological part based on the part estimation probability acquired by the part estimation probability acquisition unit 5, and controls the execution of the collation processing in the biometric information template collation unit 8. have.

  The biometric information template collation unit 8 is based on the resource allocation instructed by the priority collation control unit 7, the biometric information acquired by the biometric information acquisition unit 3, and the biometric information stored in advance in the biometric information template data holding unit 9 ( It has a function of performing verification processing with registered biometric information) and outputting an authentication result. In the biometric information template data holding unit 9, biometric information template data for a plurality of biometric parts of a plurality of authentication subjects is registered in advance by the external biometric information registration device 10 or the like. The biometric information registration device 10 has the same functions as the biometric imaging sensor 21 and the biometric information acquisition unit 3 of the biometric authentication device 1, and inputs a user ID and a biometric part using a keyboard or the like, and associates them with biometric information (template). Management and registration in the biometric information template data holding unit 9. An example of the biometric information template data holding unit 9 will be described later.

<Operation>
FIG. 6 is a flowchart showing an example of processing according to the above-described embodiment. FIG. 6A shows pre-processing, and FIG. 6B shows authentication processing.

  In FIG. 6A, when the pre-processing is started (step S1), the biometric information registration apparatus 10 performs a process on the biometric information template data holding unit 9 of the biometric authentication apparatus 1 for a plurality of biometric parts of the authentication target person. Biometric information template data is registered (step S2), and the process ends (step S3). If pre-processing has already been performed and there is no change in the person to be authenticated or the biometric information, it is not necessary to perform pre-processing repeatedly. Even in the case of performing pre-processing, only the difference from the already registered content needs to be updated.

  FIG. 7 is a diagram showing an example of biometric information template data held in the biometric information template data holding unit 9, and is an example in the case where the target biological part is the palm of the right hand and the palm of the left hand. The biometric information template data includes items of “user ID”, “biological part”, and “template”. “User ID” is an identifier indicating a subject of biometric information. “Biological part” is information for specifying a biological part. Here, the palm of the right hand is marked as “right hand”, and the palm of the left hand is marked as “left hand”, but the internal data expression can be any symbol. The “template” is binary image data indicating biometric information (biometric feature data) or coordinate value data of feature points (pattern apex, branch point, end point, etc.). Note that if the “template” includes information associated with the living body part, the living body part can be easily distinguished by processing using a program or the like, and thus the “biological part” item does not need to be explicitly provided.

  Hereinafter, the authentication process will be described.

  In FIG. 6B, when the authentication process is started by presenting the living body part to the living body sensor unit 2 (step S11), the living body imaging sensor 21 takes an image of the living body part, and the living body information acquiring unit 3 captures 2 of the captured image. Biometric information is acquired by processing such as value conversion and feature extraction (step S12). The biometric information is data in the same format as the data held as the “template” in FIG.

  Next, returning to FIG. 6B, the feature quantity acquisition unit 4 performs image processing on the image captured by the living body imaging sensor 21, acquires the feature quantity by the feature quantity sensor 22, and continues or By performing the processing, the feature amount is acquired (step S13).

  FIGS. 8A and 8B are diagrams showing an example of acquiring an angle within the photographing plane of the palm by image processing. FIGS. 8A and 8B use the contour image of the hand, and FIGS. A vein image is used. That is, as shown in FIGS. 8 (a) and 8 (b), even if the captured image has only a part of the hand and does not include the thumb that is the decisive factor for identifying left and right, part of the finger and wrist Can be detected from the direction in which the outer edge of the hand and the gap between the fingers are continuous. Further, as shown in FIGS. 8C and 8D, since many veins are directed from the base of the hand toward the fingertip, the angle of the hand in the imaging plane is detected from the average value of the direction in which the vein travels. be able to. An image such as a heel can be used instead of the vein. Since the heel is formed by grasping the hand, it is perpendicular to the direction of the vein.

  In addition, since the angle as the feature amount may be determined probabilistically, the certainty is not required. If an attempt is made to uniquely determine a living body part, since highly accurate information is required, heavy processing is required. However, since certainty is not required, light processing is sufficient. For example, it is very difficult to reliably determine left and right from an image of a hand that does not show a thumb, but it can be obtained by simple image processing if it is a rough direction.

  FIG. 9 is a diagram illustrating an example in which the angle within the photographing plane of the palm is acquired by the touch sensor as the feature amount sensor 22. That is, an arc-shaped touch sensor 22 a is provided on the fingertip side of the biological imaging sensor 21, and the detection area of the touch sensor 22 a is radially divided when viewed from the center of the biological imaging sensor 21. Therefore, when the user presents the hand to the biometric imaging sensor 21, the fingertip touches any detection area of the touch sensor 22a, and the angle of the hand in the imaging plane is detected depending on which detection area has reacted. can do.

  Although the case where image processing or the feature amount sensor 22 obtains the angle within the imaging plane of the palm is illustrated, other feature amounts can be obtained in the same manner. For example, the distance between the living body part and the living body sensor unit 2 can be an ultrasonic or optical distance sensor, and the speed can be obtained by differentiating the distance with time. A pressure sensor can be used for the pressure applied to the biosensor unit 2.

  Next, returning to FIG. 6B, the part estimation probability acquisition unit 5 refers to the feature quantity / partial estimation probability correspondence data from the feature quantity / partial estimation probability correspondence data holding unit 6 using the feature quantity as a key, and performs part estimation. A probability is acquired (step S14).

  FIG. 10 is a diagram showing an example of the feature quantity / partial estimation probability correspondence data held in the feature quantity / partial estimation probability correspondence data holding unit 6, and shows an example in which the object is the palm of the right hand and the palm of the left hand. is there. FIG. 10A shows the relationship between the hand angle and the part estimation probability (probability of being the right hand here) in a continuous graph. This graph is obtained by preliminary experiments or the like in advance when the user performs various ways of presenting the left and right palms to the biometric sensor unit 2 in the corresponding biometric authentication device 1, This is a graph of the probability of being a hand. There is a very high correlation between the angle of the hand presented to the biosensor unit 2 and whether the hand is the left or right hand, and there is a high probability that the hand is the right hand on the left side of the graph, and the left hand on the right side. There is a certain probability.

  FIGS. 10B and 10C show examples of how to provide feature quantity / region estimation probability correspondence data. In FIG. 10B, feature amount / part estimation probability correspondence data is held by a table in which angles and part estimation probabilities are associated with each other. In FIG. 10 (c), the feature amount / part estimation probability correspondence data is held by a function formula using the angle as a variable. In this case, the part estimation probability is obtained (calculated) by performing an operation by applying an angle to the function f.

  Next, returning to FIG. 6B, the priority collation control unit 7 determines the resource allocation for the collation processing for each living body part based on the part estimation probability acquired by the part estimation probability acquisition unit 5 (step S <b> 15). The resource allocation is determined by applying the part estimation probability and the verification processing time to a function having the part estimation probability and the verification processing time as variables.

  FIG. 11 is a diagram showing an example of resource allocation determined by the priority collation control unit 7. FIGS. 11 (a) and 11 (b) show examples of changes in resource allocation with respect to the collation processing time t. FIG. 11 (a) is a linear change, and FIG. 11 (b) is changed stepwise. This is the case. The method of changing is arbitrary, and the method is arbitrary as long as it satisfies the point of allocating resources based on the part estimation probability. In the example of FIGS. 11A and 11B, a resource ratio corresponding to the part estimation probability acquired by the part estimation probability acquisition unit 5 is assigned to each biological part at the start of the matching process, and until the end of the matching process. In addition, continuous resource allocation is made so that the collation processing for each biological part is completed. In addition, in order to perform verification processing for any living body part by simply changing the resource allocation, the upper and lower areas delimited by the resource allocation straight lines and curves (the lower area is the x-axis, the upper area is delimited by the 100% line) The area is proportional to the number of data of each biological information, and when the number is the same as the right hand and the left hand, the area is the same. Therefore, the time T until the collation process ends is substantially constant regardless of resource allocation.

  FIG. 11C shows an example of a function for determining resource allocation, and resource allocation is calculated by applying the site estimation probability and the matching processing time to a function g having the site estimation probability and the matching processing time as variables. .

  Next, returning to FIG. 6B, the priority collation control unit 7 controls the execution of collation processing in the biometric information template collation unit 8 based on the determined resource allocation (priority collation control) (step S16).

  FIG. 12 is a flowchart showing an example of collation processing considering resource allocation, and FIG. 12A shows the number of registered biometric information used for collation for each biological part when collation processing is performed in a single process. An example in which resource allocation is reflected, FIG. 12B is an example in which resource allocation is reflected in the number of processes allocated to each living body part when collation processing is performed in a plurality of processes.

  In FIG. 12A, when the biometric information template matching unit 8 starts processing (step S21), it acquires resource allocation from the priority matching control unit 7 (step S22).

  Next, the biometric information template data is extracted from the biometric information template data holding unit 9 at a rate based on the acquired resource allocation, and the similarity based on the verification biometric information and the template is calculated (step S23).

  Next, it is determined whether there is any remaining biometric information template data (step S24), and if it still exists (No in step S24), the process returns to acquisition of resource allocation (step S22).

  If there is no remaining biometric information template data (Yes in step S24), a user ID exceeding the predetermined threshold and having the highest similarity is output (step S25). When there is no data exceeding the predetermined threshold, “not applicable” is output. Thereafter, the process is terminated (step S26).

  In FIG.12 (b), if the biometric information template collation part 8 starts a process (step S31), it will acquire resource allocation from the priority collation control part 7 (step S32).

  Next, the number of processes is allocated at a rate based on the acquired resource allocation (step S33).

  Next, the biometric information template data is extracted for each process, and the similarity based on the verification biometric information and the template is calculated (step S34).

  Next, it is determined whether there is any remaining biometric information template data in all processes (step S35). If it still exists (No in step S35), the process returns to acquisition of resource allocation (step S32).

  If there is no remaining biometric template data in all processes (Yes in step S35), a user ID exceeding the predetermined threshold and having the highest similarity is output (step S36). When there is no data exceeding the predetermined threshold, “not applicable” is output. Thereafter, the process is terminated (step S37).

  In either case of FIGS. 12 (a) and 12 (b), there is no information as in the prior art by preferentially executing the matching process on the registration data of the biological part having a high part estimation probability. It can be expected that the registered data matching the biometric features will be reached earlier than the case where the matching process is uniformly executed for all the registered data.

  Next, returning to FIG. 6B, the biometric information template matching unit 8 outputs the authentication result when the matching process is completed (step S17), and ends the process (step S18).

<Contrast with known examples>
The known example disclosed in Patent Document 1 described above is characterized by identifying and collating which living body part the living body presented to the sensor is. However, this embodiment differs from this in that it does not uniquely determine the living body part, but obtains probabilistic information from the behavior tendency of the entire user, and uses the probability value to perform calculation processing. Allocate resources continuously.

  In addition, the present embodiment is characterized in that it utilizes the difference in the way of presentation to a living body sensor that naturally and inevitably occurs due to the limitation of the physical movable range of the human body. The known examples do not have such a feature. A relatively heavy process is required to uniquely determine a living body part as in a known example, but a feature amount such as an angle, a distance, a pressure, or the like is acquired as in this embodiment, or a probability value corresponding to it is obtained. The process to calculate can be implemented by a light process.

  In addition, when a biological part is uniquely determined, recovery in the case where the biological part is wrong becomes a form in which collation processing of another biological part is started after all the biological part verifications that have been initially determined are completed. However, in the present embodiment, since only the distribution of the verification processing resource is changed based on the probability, the verification is started at a relatively early stage as compared with the known example even for the biological part having the lower probability. For example, in the simple example, when both the left hand and right hand probabilities are 50%, in the known example, one hand is decided once and the collation process is completed before the other hand collation process. Therefore, although one of the probabilities is equal, the processing of either hand is postponed, and there is a possibility that the time for the collation processing becomes longer. In the present embodiment, if the probability is 50%, the same 50% verification processing resources are allocated to the verification processing of both hands, so there is no demerit that the verification processing of either hand is delayed. Therefore, this embodiment is more advantageous than the known example in the case of a delicate probability of 60% vs. 40%.

<Summary>
As described above, the present embodiment has the following advantages.
(1) There is no need to present a personal ID or the like in advance, and by leaving the merit of one-to-N authentication that performs personal authentication simply by holding a biological part over a biometric sensor, no extra burden is imposed on the user. Nevertheless, it is possible to narrow down the registration data by using supplementary information such as the tendency when the biometric authentication device is used by the entire user, thereby shortening the verification processing time. It becomes. That is, convenience is improved without increasing the burden on the user.
(2) In the conventional biometric authentication method, it is possible to improve the authentication speed by effectively utilizing information that has not been effectively utilized as being unnecessary to confirm the identity.
(3) When processing is performed using an image captured by the biosensor, there is almost no need to change the existing hardware, and only a software change is required, so that the cost effectiveness is great.

The present invention has been described above by the preferred embodiments of the present invention. While the invention has been described with reference to specific embodiments, various modifications and changes may be made to the embodiments without departing from the broad spirit and scope of the invention as defined in the claims. Obviously you can. In other words, the present invention should not be construed as being limited by the details of the specific examples and the accompanying drawings.
(Appendix 1)
Biometric information holding means for holding biometric information about a plurality of biometric parts of a plurality of authentication subjects as registered biometric information so that the authentication target person and the biometric part can be specified;
Biometric information acquisition means for acquiring biometric information of a user as authentication biometric information at the time of authentication;
Feature amount acquisition means for acquiring a feature amount indicating a method of presenting a biological part at the time of authentication;
A site estimation probability acquisition means for acquiring a site estimation probability indicating the probability of being a predetermined biological site based on the feature amount;
Collation control means for determining the resource allocation of the collation process for each living body part based on the part estimation probability and controlling execution of the collation process;
A biometric authentication apparatus comprising: a verification unit that performs a verification process of the authentication biometric information and the registered biometric information based on the resource allocation.
(Appendix 2)
In the biometric authentication device according to attachment 1,
The biometric authentication apparatus, wherein the feature quantity acquisition unit acquires a feature quantity by performing image processing on an image acquired from the biological part by the biometric information acquisition unit.
(Appendix 3)
In the biometric authentication device according to attachment 1,
The biometric authentication apparatus, wherein the feature quantity acquisition means acquires a feature quantity by a feature quantity sensor.
(Appendix 4)
In the biometric authentication device according to any one of appendices 1 to 3,
The part estimation probability acquisition unit acquires a corresponding part estimation probability from a table in which feature quantities and part estimation probabilities are associated in advance with the acquired feature quantity as a key.
(Appendix 5)
In the biometric authentication device according to any one of appendices 1 to 3,
The part estimation probability acquisition unit calculates a part estimation probability by applying a feature amount to a function in which a feature amount and a part estimation probability are associated in advance.
(Appendix 6)
In the biometric authentication device according to any one of appendices 1 to 5,
The collation control means assigns a resource ratio according to the acquired part estimation probability to each biological part at the start of the collation process, and the collation process for each biological part is completed by the end of the collation process. A biometric authentication apparatus characterized by determining an appropriate resource allocation.
(Appendix 7)
In the biometric authentication device according to any one of appendices 1 to 6,
The verification unit extracts the registered biometric information of each biological part at a rate based on the resource allocation at each time point indicated by the verification control unit as the verification process progresses, and performs the verification process apparatus.
(Appendix 8)
In the biometric authentication device according to any one of appendices 1 to 6,
The biometric authentication apparatus, wherein the collating unit performs collation processing of each biological part by a process with a ratio of the number of processes based on resource allocation at each time pointed by the collating control unit as time passes in the collating process .
(Appendix 9)
In the biometric authentication device according to any one of appendices 1 to 8,
The biometric authentication apparatus, wherein the feature amount is an angle of a biological part within a photographing surface of biometric information.
(Appendix 10)
In the biometric authentication device according to any one of appendices 1 to 8,
The biometric authentication apparatus, wherein the feature amount is an angle of a biological part with respect to a photographing surface of biometric information.
(Appendix 11)
In the biometric authentication device according to any one of appendices 1 to 8,
The biometric authentication device, wherein the feature amount is a distance of a biological part with respect to a photographing surface of biological information.
(Appendix 12)
In the biometric authentication device according to any one of appendices 1 to 8,
The biometric authentication apparatus, wherein the feature amount is a speed of a biological part relative to a photographing surface of biometric information.
(Appendix 13)
In the biometric authentication device according to any one of appendices 1 to 8,
The biometric authentication device, wherein the feature amount is a pressure of a biological part with respect to a photographing surface of biometric information.
(Appendix 14)
Registering biometric information about a plurality of biometric parts of a plurality of authentication subjects as registered biometric information so that the authentication target person and the biometric part can be identified;
Acquiring the user's biometric information as authentication biometric information at the time of authentication;
Obtaining a feature amount indicating how to present a living body part at the time of authentication;
Obtaining a site estimation probability indicating the probability of being a predetermined biological site based on the feature amount;
Determining the resource allocation of the matching process for each living body part based on the part estimation probability and controlling the execution of the matching process;
A biometric authentication method comprising a step of performing a verification process of the authentication biometric information and the registered biometric information based on the resource allocation.

DESCRIPTION OF SYMBOLS 1 Biometric authentication apparatus 2, 2A, 2B Biometric sensor part 21 Biometric imaging sensor 22 Feature-value sensor 22a Touch sensor 3 Biometric information acquisition part 4 Feature-value acquisition part 5 Site estimation probability acquisition part 6 Feature-value / site-estimation probability correspondence data holding part 7 priority collation control unit 8 biometric information template collation unit 9 biometric information template data holding unit 10 biometric information registration device U1, U2 user

Claims (6)

Biometric information holding means for holding biometric information about a plurality of biometric parts of a plurality of authentication subjects as registered biometric information so that the authentication target person and the biometric part can be specified;
Biometric information acquisition means for acquiring biometric information of a user as authentication biometric information at the time of authentication;
Feature amount acquisition means for acquiring a feature amount indicating a method of presenting a biological part at the time of authentication;
A site estimation probability acquisition means for acquiring a site estimation probability indicating the probability of being a predetermined biological site based on the feature amount;
Collation control means for determining the resource allocation of the collation process for each living body part based on the part estimation probability and controlling execution of the collation process;
A verification unit that performs verification processing of the authentication biometric information and the registered biometric information based on the resource allocation ,
The collation control means assigns a resource ratio according to the acquired part estimation probability to each biological part at the start of the collation process, and the collation process for each biological part is completed by the end of the collation process. A biometric authentication apparatus characterized by determining a proper resource allocation . The biometric authentication device according to claim 1,
The biometric authentication apparatus, wherein the feature quantity acquisition unit acquires a feature quantity by performing image processing on an image acquired from the biological part by the biometric information acquisition unit. The biometric authentication device according to claim 1,
The biometric authentication apparatus, wherein the feature quantity acquisition means acquires a feature quantity by a feature quantity sensor. The biometric authentication device according to any one of claims 1 to 3 ,
The verification unit extracts the registered biometric information of each biological part at a rate based on the resource allocation at each time point indicated by the verification control unit as the verification process progresses, and performs the verification process apparatus. The biometric authentication device according to any one of claims 1 to 3 ,
The biometric authentication apparatus, wherein the collating unit performs collation processing of each biological part by a process with a ratio of the number of processes based on resource allocation at each time pointed by the collating control unit as time passes in the collating process . Registering biometric information about a plurality of biometric parts of a plurality of authentication subjects as registered biometric information so that the authentication target person and the biometric part can be identified;
Acquiring the user's biometric information as authentication biometric information at the time of authentication;
Obtaining a feature amount indicating how to present a living body part at the time of authentication;
Obtaining a site estimation probability indicating the probability of being a predetermined biological site based on the feature amount;
Determining the resource allocation of the matching process for each living body part based on the part estimation probability and controlling the execution of the matching process;
The verification biometric information and the registered biometric information matching process based on the resource allocation ,
The step of controlling the execution of the collation process assigns a resource ratio corresponding to the acquired part estimation probability to each biological part at the start of the collation process, and the collation process for each biological part is completed by the end of the collation process. A biometric authentication method characterized by determining a continuous resource allocation to be terminated .

Images