Authentication device and authentication method

Abstract

To provide an authentication device capable of achieving highly accurate authentication that takes into account quality of biological information used for identity verification authentication.SOLUTION: An authentication device includes: an acquisition unit that acquires at least first biological information of an authentication target used for first biometric authentication and first surrounding environment information when acquiring the first biological information; a score calculation unit that uses first biological information to calculate a score of the first biometric authentication; a reliability determination unit that determines reliability of the first biological information from first surrounding environment information; a continuation processing determination unit that determines whether to perform second biometric authentication that uses, second biological information different from the first biological information on the basis of the score of the first biometric authentication and the reliability of the first biological information.SELECTED DRAWING: Figure 5

Description

Technical Field

[0001] The present disclosure relates to an authentication device and an authentication method.

Background Art

[0002] Patent Document 1 discloses a personal authentication system configured to perform personal authentication in response to an authentication request from a server. The personal authentication system has an authentication means that performs authentication using biometric information and has a function of outputting a score that is the similarity of the biometric information or a hash value thereof. The personal authentication system has a score log storage unit that stores a log of the score of personal authentication obtained by the authentication means or a hash value thereof. The personal authentication system has a processing unit that determines that authentication fails when the same score continues for a plurality of times based on the score or the log of the hash value read from the score log storage unit.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In recent years, there is a personal authentication technology that uses biometric information to authenticate whether a person to be authenticated matches a pre-registered person. In personal authentication technology, it is expected to improve security against impersonation and the like by performing authentication using a plurality of types (multimodal) of biometric authentication means (hereinafter referred to as multimodal authentication).

[0005] In the personal authentication system described in Patent Document 1, if the score obtained when performing authentication using a certain biometric authentication method falls below a predetermined minimum level, the processing unit stops processing without proceeding to the next authentication process. However, depending on the environment in which the biometric information is acquired, the score may be lower than when acquired under optimal conditions. For example, when performing authentication from an image of the face of the person to be authenticated captured by a camera, if the surrounding environment of the person to be authenticated is dark, the face in the captured image may become unclear, resulting in a lower score. Thus, if biometric information is acquired in an inappropriate environment, the score may be low, authentication may be terminated, and multimodal authentication may not be performed properly. As a result, conventional multimodal authentication has had the problem of misauthentication occurring, where the person to be authenticated is judged not to be the real person even if they are, thus reducing the authentication accuracy.

[0006] This disclosure was devised in light of the conventional circumstances described above, and aims to achieve highly accurate authentication that takes into account the quality of biometric information used for identity verification. [Means for solving the problem]

[0007] This disclosure includes: an acquisition unit that acquires at least first biometric information of an authentication target to be used in first biometric authentication and first surrounding environment information at the time of acquisition of the first biometric information; a score calculation unit that calculates a score for the first biometric authentication using the first biometric information; a reliability determination unit that determines the reliability of the first biometric information from the first surrounding environment information; and a continuation processing determination unit that determines whether or not to perform a second biometric authentication using second biometric information different from the first biometric information, based on the score of the first biometric authentication and the reliability of the first biometric information. The continuation processing determination unit determines that if the score of the first biometric authentication is greater than or equal to the first threshold but less than the second threshold, or if the score of the first biometric authentication is less than the first threshold and the confidence level of the first biometric information is less than the third threshold, the authentication result of the authentication target is determined to be the third determination, and the second biometric authentication is to be performed further. We provide authentication devices. Furthermore, this disclosure provides an authentication device comprising: an acquisition unit that acquires at least first biometric information of an authentication target to be used in first biometric authentication and first surrounding environment information at the time of acquisition of the first biometric information; a score calculation unit that calculates a score for the first biometric authentication using the first biometric information; a reliability determination unit that determines the reliability of the first biometric information from the first surrounding environment information; and a continuation processing determination unit that determines whether or not to perform a second biometric authentication using second biometric information different from the first biometric information, based on the score of the first biometric authentication and the reliability of the first biometric information, wherein the continuation processing determination unit determines that if the score of the first biometric authentication is equal to or greater than a second threshold and the reliability of the first biometric information is less than a fourth threshold, the authentication result of the authentication target is a fourth determination and that the second biometric authentication is to be performed further; and if the score of the second biometric authentication is equal to or greater than a second threshold, the authentication target is determined to be the real person. Furthermore, this disclosure provides an authentication device comprising: an acquisition unit that acquires at least first biometric information of an authentication target to be used in first biometric authentication and first ambient environment information at the time of acquisition of the first biometric information; a score calculation unit that calculates a score for the first biometric authentication using the first biometric information; a reliability determination unit that determines the reliability of the first biometric information from the first ambient environment information; and a continuation processing determination unit that determines whether or not to perform a second biometric authentication using second biometric information different from the first biometric information, wherein the continuation processing determination unit determines that if the score of the first biometric authentication is greater than or equal to a first threshold and less than a second threshold, or if the score of the first biometric authentication is less than a first threshold and the reliability of the first biometric information is less than a third threshold, the authentication result of the authentication target is determined to be a third determination; if the score of the first biometric authentication is greater than or equal to a second threshold and the reliability of the first biometric information is less than a fourth threshold, the authentication result of the authentication target is determined to be a fourth determination; and in the case of the third determination and the fourth determination, it determines that the second biometric authentication will be performed further.

[0008] Furthermore, this disclosure relates to an authentication method performed by an authentication device that authenticates the identity of a person to be authenticated, comprising: acquiring at least the first biometric information of the person to be authenticated and the first surrounding environment information at the time the first biometric information was acquired; calculating a score for the first biometric authentication using the first biometric information; determining the reliability of the first biometric information from the first surrounding environment information; and determining whether or not to perform a second biometric authentication using second biometric information different from the first biometric information, based on the score for the first biometric authentication and the reliability of the first biometric information. death , If the score of the first biometric authentication is greater than or equal to the first threshold but less than the second threshold, or if the score of the first biometric authentication is less than the first threshold and the confidence level of the first biometric information is less than the third threshold, the authentication result of the authentication target is determined to be the third determination, and it is determined that the second biometric authentication should be performed further. Provide an authentication method. Furthermore, this disclosure provides an authentication method performed by an authentication device that authenticates the identity of a person to be authenticated, comprising: acquiring at least first biometric information of the person to be authenticated and first surrounding environment information at the time of acquisition of the first biometric information used in first biometric authentication; calculating a score for the first biometric authentication using the first biometric information; determining the reliability of the first biometric information from the first surrounding environment information; determining whether or not to perform a second biometric authentication using second biometric information different from the first biometric information based on the score of the first biometric authentication and the reliability of the first biometric information; if the score of the first biometric authentication is equal to or greater than a second threshold and the reliability of the first biometric information is less than a fourth threshold, the authentication result of the person to be authenticated is determined to be a fourth determination, and it is determined that the second biometric authentication will be performed further; and if the score of the second biometric authentication is equal to or greater than a second threshold, it is determined that the person to be authenticated is the person in question. Furthermore, this disclosure provides an authentication method performed by an authentication device that authenticates the identity of an object to be authenticated, comprising: acquiring at least first biometric information of the object to be authenticated and first ambient environment information at the time of acquisition of the first biometric information used in first biometric authentication; calculating a score for the first biometric authentication using the first biometric information; determining the reliability of the first biometric information from the first ambient environment information; determining whether or not to perform a second biometric authentication using second biometric information different from the first biometric information based on the score of the first biometric authentication and the reliability of the first biometric information; determining the authentication result of the object to be authenticated as a third determination if the score of the first biometric authentication is above a first threshold and below a second threshold, or if the score of the first biometric authentication is below a first threshold and the reliability of the first biometric information is below a third threshold; determining the authentication result of the object to be authenticated as a fourth determination if the score of the first biometric authentication is above a second threshold and the reliability of the first biometric information is below a fourth threshold; and determining whether or not to further perform the second biometric authentication in the case of the third determination and the fourth determination.

[0009] These comprehensive or specific embodiments may be implemented as systems, devices, methods, integrated circuits, computer programs, or recording media, or as any combination of systems, devices, methods, integrated circuits, computer programs, and recording media. [Effects of the Invention]

[0010] According to this disclosure, it is possible to achieve highly accurate authentication that takes into account the quality of biometric information used for identity verification. [Brief explanation of the drawing]

[0011] [Figure 1] A diagram illustrating an example of identity verification authentication. [Figure 2] A diagram showing the hardware configuration of the authentication system according to this embodiment. [Figure 3] Diagram showing the internal configuration of the processor according to this embodiment. [Figure 4] A diagram showing an example of case classification of the determination result according to this embodiment. [Figure 5] Flowchart showing the continuous processing of biometric authentication according to Embodiment 1 [Figure 6] Conceptual diagram of a case where a second biometric authentication method is selected based on the first biometric authentication method. [Figure 7]Figure showing an example of an authentication method when the reliability of all biometric authentications is low [Figure 8] Figure representing an example in which it is determined that the authentication target person is not the actual person based on the results of multiple biometric authentications [Figure 9] Flowchart of the continuous biometric authentication process according to Embodiment 2 [Figure 10] Figure showing an example of determining the first biometric authentication based on the information of the surrounding environment acquired before the start of authentication [Figure 11] Figure showing an example of the reliability of each biometric information

Embodiments for Carrying Out the Invention

[0012] Hereinafter, with appropriate reference to the drawings, embodiments specifically disclosing the authentication apparatus and authentication method according to the present disclosure will be described in detail. However, detailed descriptions that are more than necessary may be omitted. For example, detailed descriptions of well-known matters and duplicate descriptions of substantially the same configurations may be omitted. This is to avoid making the following description unnecessarily redundant and to facilitate the understanding of those skilled in the art. The attached drawings and the following description are provided for those skilled in the art to fully understand the present disclosure, and are not intended to limit the subject matter described in the claims.

[0013] <Summary of the Invention> The authentication device 10 (see FIG. 2) according to this embodiment performs authentication of verifying the identity of the authentication target by comparing the pre-registered biometric information with the biometric information acquired by various sensors. The authentication device 10 performs multi-modal authentication based on a plurality of biometric information. The authentication device 10 compares the acquired biometric information of the authentication target with the pre-registered biometric information and calculates the matching rate as a score. However, the quality of the biometric information may deteriorate depending on the environment in which the biometric information is acquired, and the accuracy of authentication may deteriorate. Also, multi-modal authentication takes time to acquire a plurality of biometric information, and the burden on the authentication target may increase. The authentication device 10 estimates the quality of the biometric information as well as the score, and calculates the reliability of the score. By integrating a plurality of modalities in consideration of the reliability of this score, highly accurate and efficient authentication is realized. Hereinafter, an index calculated based on the quality of biometric information (which can also be said to be an index indicating whether the data of the biometric information used for biometric authentication is reliable) is referred to as reliability. For example, the acquired biometric information is compared with the pre-registered biometric information to calculate the score and the reliability of the score. Generally, biometric information acquired from sensors such as cameras or microphones includes noise unrelated to individuality in addition to information representing the individuality of the authentication target. For example, a face image includes noise included in the image itself, or accessories such as glasses, sunglasses, or masks that have no direct relation to individuality as noise unrelated to individuality. Voice includes background noise such as environmental noise as noise unrelated to individuality. When the noise component is relatively large with respect to the component representing the individuality of the authentication target, it is presumed that the quality of the biometric information is low. When the quality of this biometric information is low, the score obtained from the biometric information has low reliability. This reliability can be estimated based on the brightness of the image or the background noise level outside the section where the authentication target is speaking.

[0014] Referring to FIG. 1, the outline of the authentication determination for identity verification will be described. FIG. 1 is a diagram showing an example of the authentication determination for identity verification.

[0015] Case CA illustrates an example where the person being authenticated was determined not to be the person they claim to be. Hereafter, a result of biometric authentication determining that the person being authenticated is not the person they claim to be will be referred to as an "NG" (Not Accepted) judgment.

[0016] Authentication Cer1 shows the result of biometric authentication, which involves acquiring biometric information related to the face of the person being authenticated and comparing it with the person's pre-stored biometric information. The score takes a value from 0 to 100, for example, with a higher number indicating a higher score. In other words, a higher number indicates a higher biometric authentication result. The confidence value takes a value similar to the score. In other words, a higher number indicates higher confidence, and indicates that the environment in which the biometric information was acquired was more suitable for the execution of biometric authentication using that biometric information. Note that the values ​​taken for the score and confidence value are examples and are not limited to 0 to 100. The score for Authentication Cer1 is a low value of 20. Authentication Cer1 acquired biometric information related to the face of the person being authenticated in an environment that was bright enough to clearly identify the face, so the confidence value of the score is a high value of 100. As a result, Case CA judges it as NG based on the score and confidence value of Authentication Cer1.

[0017] Case CB illustrates an example where the person being authenticated was determined to be the person in question. Hereafter, when the biometric authentication result determines that the person being authenticated is the person in question, this will be referred to as an "OK" judgment.

[0018] Authentication Cer2 shows the result of biometric authentication, which involves acquiring biometric information related to the face of the person being authenticated and comparing it with the person's pre-stored biometric information. The score for Authentication Cer2 is a low value of 50. However, in Authentication Cer2, the biometric information related to the face of the person being authenticated was acquired in a dark environment where the face was unclear, resulting in a low confidence score of 50. As a result, the confidence score of the biometric information from Authentication Cer2 is low, and it is determined that additional authentication is required.

[0019] Authentication Cer3 is the result of biometric authentication performed after Authentication Cer2. Authentication Cer3 captures the speech of the person being authenticated, which is different biometric information from the facial biometric information used in Authentication Cer2, which was performed beforehand. The score for Authentication Cer3 is a high value of 90. Authentication Cer3 has a high confidence score of 100 because the biometric information related to the speech was acquired in a quiet environment where the speech of the person being authenticated could be sufficiently determined.

[0020] Based on the results of Cer2 and Cer3, Case CB is judged as OK.

[0021] Hereafter, the first biometric authentication (Authentication Cer2 in Case CB) will be referred to as the first biometric authentication, and the second biometric authentication (Authentication Cer3 in Case CB) will be referred to as the second biometric authentication. Note that the number of authentications is not limited to two, and may be two or more, but for the sake of explanation, the terms first biometric authentication and second biometric authentication will be used. The biometric information used in the first biometric authentication will be referred to as the first biometric information, and the biometric information used in the second biometric authentication will be referred to as the second biometric information.

[0022] Hereinafter, information related to the external environment obtained through the first biometric authentication will be referred to as the first environmental information, and information related to the external environment obtained through the second biometric authentication will be referred to as the second environmental information.

[0023] Thus, the authentication device 10 according to this embodiment performs authentication for verifying the identity of the person to be authenticated by integrally determining the score and confidence level of the person's biometric information.

[0024] <Hardware configuration of this embodiment> Next, the hardware configuration of the authentication system according to this embodiment will be described with reference to Figure 2. Figure 2 is a diagram showing the hardware configuration of the authentication system according to this embodiment.

[0025] The authentication system 1 includes an authentication device 10, a user database 30, and a drive device 40. Here, DB stands for database. Note that the user database 30 and the drive device 40 may be included in the authentication device 10.

[0026] The authentication device 10 includes a sensor 20, a display 21, a processor 22, a memory 23, and a communication IF 25. Here, IF stands for interface. The authentication device 10 is a housing or the like equipped with a sensor 20 related to biometric authentication. The authentication device 10 may be a tablet, a PC (Personal Computer), a mobile terminal, etc. Furthermore, the authentication device 10 may be equipped with multiple sensors in a single device, or it may consist of multiple devices, each equipped with at least one sensor. The authentication device 10 may be installed, for example, at the entrance of a commercial facility, amusement park, tourist attraction, or building, and used to verify the identity of visitors. Alternatively, the authentication device 10 may be installed in an ATM (Automatic Teller Machine) and used to verify the identity of users. Note that the installation locations of the authentication device 10 are examples only and are not limited to these.

[0027] The sensor 20 includes a face camera 201, a fingerprint camera 202, and a microphone 203. The face camera 201 and the fingerprint camera 202 may use the same camera.

[0028] The face camera 201 is a camera that captures images of the face of the person to be authenticated. The face camera 201 is composed of at least a lens (not shown) and an image sensor (not shown) as optical elements. The lens receives light reflected from an object within the field of view of the area captured by the face camera 201 and forms an optical image of the object on the light-receiving surface (in other words, the imaging surface) of the image sensor. The image sensor is a solid-state image sensor such as a CCD (Charged Coupled Device) or CMOS (Complementary Metal Oxide Semiconductor). At predetermined time intervals (e.g., 1 / 30 (second)), the image sensor converts the optical image formed on the imaging surface via the lens into an electrical signal and sends it to the processor 22. For example, if the predetermined time interval is 1 / 30 (second), the frame rate of the face camera 201 will be 30 fps. The face camera 201 may also generate image data (video data) by applying predetermined signal processing to the electrical signal at the predetermined time intervals mentioned above. The process of generating this image data (video data) may be performed by the processor 22. The face camera 201 outputs the image data (video data) to the processor 22. Hereinafter, image data or video data will be referred to as imaging data.

[0029] The fingerprint camera 202 is a camera that captures the fingerprints of a person to be authenticated. The fingerprint camera 202 is configured to have at least a lens (not shown) and an image sensor (not shown) as optical elements. The lens receives light reflected from an object within the field of view of the area captured by the fingerprint camera 202 and forms an optical image of the object on the light-receiving surface (in other words, the imaging surface) of the image sensor. The image sensor is a solid-state image sensor such as a CCD or CMOS. At predetermined intervals (e.g., 1 / 30 (second)), the image sensor converts the optical image formed on the imaging surface via the lens into an electrical signal and sends it to the processor 22. The fingerprint camera 202 may also generate imaging data by applying predetermined signal processing to the electrical signal at the aforementioned predetermined intervals. Note that this processing to generate imaging data may be performed by the processor 22. The fingerprint camera 202 outputs the imaging data to the processor 22.

[0030] The face camera 201 and the fingerprint camera 202 may acquire information about the surrounding environment of the authentication device 10 (hereinafter referred to as "surrounding environment information"). The surrounding environment information acquired by the face camera 201 or the fingerprint camera 202 may include, for example, brightness. Note that the acquisition of surrounding environment information is not limited to the face camera 201 and the fingerprint camera 202, but may also be acquired from sensors including other cameras provided in or outside the authentication device 10.

[0031] Microphone 203 captures the speech of the person being authenticated. Microphone 203 consists of a microphone device capable of capturing (i.e., detecting) the sound generated based on the speech of the person being authenticated. Microphone 203 captures the sound generated based on the speech of the person being authenticated, converts it into an electrical signal as captured data, and outputs it to processor 22.

[0032] Microphone 203 may acquire ambient environmental information. Ambient environmental information includes, for example, noise other than the voice of the person being authenticated. Note that the acquisition of ambient environmental information is not limited to microphone 203, but may also be acquired from sensors including other microphones provided on or outside the authentication device 10.

[0033] The display 21 displays the authentication result obtained from the processor 22. Note that the display 21 may be omitted from the configuration of the authentication device 10. For example, when opening a door using biometric authentication, the authentication result can be determined by the opening and closing of the door, so displaying the authentication result on the display 21 is not necessarily required.

[0034] The processor 22 uses a semiconductor chip on which at least one of the following electronic devices is implemented: a CPU (Central Processing Unit), a DSP (Digital Signal Processor), a GPU (Graphical Processing Unit), or an FPGA (Field Programmable Gate Array). The processor 22 functions as a controller that oversees the overall operation of the authentication device 10, performing control processing to coordinate the operation of each part of the authentication device 10, data input / output processing between each part of the authentication device 10, data calculation processing, and data storage processing.

[0035] The processor 22 uses programs and data stored in the ROM (Read Only Memory) of the memory 23 to implement the functions of the overall control unit 50 (see Figure 3), the modal processing unit 60 (see Figure 3), and the modal processing unit 70 (see Figure 3). During operation, the processor 22 uses the RAM (Random Access Memory) of the memory 23 to temporarily store data or information generated or acquired by the processor 22 and each unit in the RAM of the memory 23.

[0036] Memory 23 includes, for example, a ROM that stores a program defining various processes performed by the processor 22 and data used during the execution of that program, and RAM as work memory used when executing various processes performed by the processor 22. The ROM 22 contains a program defining various processes performed by the processor 22 and data used during the execution of that program. The RAM temporarily stores data or information generated or acquired by the processor 22 (for example, biometric information acquired by the sensor 20, or biometric information scores and confidence levels). The ROM may be replaced with a storage medium such as flash memory, HDD (Hard Disk Drive), or SSD (Solid State Drive), or an external server device, and the data may be read when the device starts up.

[0037] The input device 24 receives input from the person to be authenticated. The input device 24 is, for example, a touch panel display or a sensor capable of authenticating an access card or employee ID. The input device 24 receives input from the person to be authenticated, such as name, gender, age, ID or password. The input device 24 may also obtain information related to the input content from the user DB 30 and perform authentication with the obtained person (i.e., one-to-one authentication). Note that the input device 24 may be omitted from the configuration of the authentication device 10.

[0038] The communication IF25 is a network interface circuit that communicates wirelessly or via wired connection between the authentication device 10 and the user DB 30 and the drive device 40, respectively. The communication methods used for communication at the communication IF25 include, for example, mobile communications such as WAN (Wide Area Network), LAN (Local Area Network), LTE (Long Term Evolution), and 5G, as well as power line communications, short-range wireless communications (e.g., Bluetooth® communications), and communications for mobile phones.

[0039] The user database 30 is configured using a storage medium (e.g., flash memory, HDD (Hard Disk Drive), or SSD (Solid State Drive)). The user database 30 stores previously acquired biometric data and various other information (name, age, gender, facial photograph, ID or password, etc.) of individuals used in pre-registered authentication processes. Biometric information includes facial image data, feature data of feature points on facial image data, fingerprint or speech data extracted from individuals in the past, feature data of feature points on fingerprint image data, and feature data of speech. However, biometric information is not limited to these. The user database 30 may be stored on a cloud server and connected to the authentication device 10 via a network.

[0040] The drive unit 40 performs actions such as opening and closing a gate installed near the authentication device 10, sounding a buzzer from a speaker mounted on the authentication device 10, or turning on or flashing a light (e.g., an LED light) mounted on the authentication device 10. However, the actions performed by the drive unit 40 are not limited to these.

[0041] <Internal configuration of the processor in this embodiment> Next, the internal configuration of the processor according to this embodiment will be described with reference to Figure 3. Figure 3 is a diagram showing the internal configuration of the processor according to this embodiment.

[0042] The processor 22 implements the functions of the overall control unit 50, the modal processing unit 60, and the modal processing unit 70. The overall control unit 50, the modal processing unit 60, and the modal processing unit 70 may be integrated as functions of a single processor, as shown in Figure 3, or they may be integrated as functions of separate processors. The number of modal processing units is not limited to two, and there may be two or more depending on the type of biometric information used for authentication.

[0043] The overall control unit 50 performs authentication processing for verifying the identity of the person to be authenticated. The overall control unit 50 obtains the score and confidence level calculated from the modal processing unit 60 and the modal processing unit 70. The overall control unit 50 includes an integrated determination unit 51, a continuation processing determination unit 52, a system setting unit 53, and a user characteristics determination unit 54.

[0044] The integrated determination unit 51 determines whether the result is OK or NG based on the determination result of the first biometric authentication obtained from the continuous processing determination unit 52, or based on the combined determination results of the first and second biometric authentications. The number of biometric authentications used by the integrated determination unit 51 may be two or more.

[0045] The integrated determination unit 51 performs processing using scores and confidence levels related to multiple biometric information (e.g., facial imaging data, fingerprint imaging data, or speech recording data). For example, the integrated determination unit 51 calculates the sum of the scores related to the multiple biometric information. The integrated determination unit 51 compares the sum of the scores with a threshold related to the sum to determine whether the person being authenticated is the real person or not. The threshold related to the sum may be held by the integrated determination unit 51 or stored in the memory 23. The integrated determination unit 51 may also calculate the integrated false acceptance rate. The false acceptance rate represents the probability of misauthentication, where a person who is not the real person is recognized as the real person when the determination of whether the person being authenticated is the real person is made using a predetermined threshold. For example, a false acceptance rate of 1% means that misauthentication occurs with a probability of 1 in 100 times. The integrated false acceptance rate is the false acceptance rate when the determination is made based on multiple biometric information, and is the value obtained by multiplying the false acceptance rates based on each biometric information and taking the reciprocal. For example, if the false acceptance rate for authentication using biometric information related to the subject's face is 1%, and the false acceptance rate for authentication using biometric information related to speech is 10%, the combined false acceptance rate will be 0.1%. Note that if the multiple biometric data used to calculate the combined false acceptance rate are not mutually independent, a correction may be made to consider the correlation between the biometric data.

[0046] The continuation processing determination unit 52 obtains the score of the first biometric authentication and the confidence level of the first biometric information calculated from the modal processing unit 60 or the modal processing unit 70. The continuation processing determination unit 52 compares the score with a threshold related to the score, and compares the confidence level with a threshold related to the confidence level, to determine whether or not to perform the second biometric authentication. The second biometric authentication is performed using biometric information different from the biometric information used in the previously performed biometric authentication. For example, if the continuation processing determination unit 52 determines that a second biometric authentication is necessary after biometric authentication has been performed using biometric information related to the face, it performs a second biometric authentication using biometric information other than the face (e.g., fingerprints or speech).

[0047] The continuation processing determination unit 52 determines, for example, whether the first biometric authentication falls under a confirmation determination, verification determination, doubt determination, or rejection determination (see Figure 4). Confirmation determination, verification determination, doubt determination, and rejection determination are described in detail in Figure 4. Based on the determined determination result, the continuation processing determination unit 52 decides whether or not to perform the second biometric authentication.

[0048] The system setting unit 53 sets the order of biometric authentication. The order of biometric authentication may be a predetermined order or an order determined by the continuation processing determination unit 52. The system setting unit 53 also sets the score and confidence thresholds. The score and confidence thresholds may be predetermined values ​​or may be set by the system setting unit 53 according to the person being authenticated or the surrounding environment information. For example, if it is known from the user DB 30 that the person being authenticated is a twin, the system setting unit 53 sets a higher threshold for biometric information related to the face. It also sets a higher threshold for biometric information of poor quality (poor image or sound quality, etc.) registered in the user DB 30. Information related to the predetermined order is held by the system setting unit 53 or memory 23. Information related to the predetermined score and confidence thresholds is held by the system setting unit 53, modal processing unit 60, modal processing unit 70, or memory 23. Hereinafter, the biometric authentication order, score thresholds, and confidence thresholds set by the system setting unit 53 will be referred to as system setting information.

[0049] The user characteristic determination unit 54 extracts the characteristics of the person to be used for biometric authentication obtained from the user database 30. Characteristics include, for example, whether the person to be authenticated is a twin. The user characteristic determination unit 54 outputs the extracted information to the system setting unit 53. The system setting unit 53 may also directly obtain the characteristics of the person to be used for biometric authentication from the user database 30.

[0050] The modal processing unit 60 performs processing using biological information or surrounding environment information acquired from the sensor 20. The modal processing unit 60 includes a reliability determination unit 61 and a score calculation unit 62.

[0051] The reliability determination unit 61 acquires surrounding environment information from the sensor 20 and determines the reliability of the biometric authentication.

[0052] The score calculation unit 62 acquires biometric information from the sensor 20, compares it with the information registered in the user database 30, and calculates a score.

[0053] The modal processing unit 70 performs processing using different biological information than the modal processing unit 60. Since the functions of the modal processing unit 70 are the same as those of the modal processing unit 60, a detailed explanation is omitted.

[0054] Hereinafter, the process of calculating confidence and scores performed by the modal processing unit 60 or modal processing unit 70 will be referred to as modal authentication.

[0055] <An example of how to categorize biometric authentication results> Next, with reference to Figure 4, an example of the case classification of the judgment result according to this embodiment will be explained. Figure 4 is a diagram showing an example of the case classification of the judgment result according to this embodiment.

[0056] In the graph shown in Figure 4, the vertical axis represents the score value, with the score increasing in the direction of the arrow. The score has a first threshold and a second threshold. Note that the number of thresholds a score has is not limited to two; it may be one or more. For example, the first threshold is 30 and the second threshold is 70. Note that the values ​​of the first and second thresholds are examples only and are not limited to these.

[0057] In the graph shown in Figure 4, the horizontal axis represents the confidence level, with confidence increasing along the direction indicated by the arrow. The confidence level has a third threshold and a fourth threshold. The number of thresholds is not limited to two; there may be one or more. For example, the third threshold is 30 and the fourth threshold is 70. The values ​​for the third and fourth thresholds are examples only and are not limited to these.

[0058] In the graph shown in Figure 4, the filled-in areas indicate that a second biometric authentication is required.

[0059] The continuation processing determination unit 52 determines that the authentication result of the first biometric authentication of the person being authenticated is a "confirmation judgment," meaning the person being authenticated is indeed the person in question, if the score of the first biometric authentication is equal to or greater than the second threshold and the confidence level of the first biometric information is equal to or greater than the fourth threshold. If the first biometric authentication is a confirmation judgment, the continuation processing determination unit 52 decides not to perform the second biometric authentication.

[0060] The continuation processing determination unit 52 determines that the authentication result of the first biometric authentication of the person being authenticated is a "rejection," meaning the person being authenticated is not the person in question, if the score of the first biometric authentication is below the first threshold and the confidence level of the first biometric information is at or above the third threshold. If the first biometric authentication is a rejection, the continuation processing determination unit 52 decides not to perform the second biometric authentication.

[0061] The continuation processing determination unit 52 determines the authentication result of the first biometric authentication of the person being authenticated as a "doubtful determination" if the score of the first biometric authentication is greater than or equal to the first threshold but less than the second threshold, or if the score of the first biometric authentication is less than the first threshold and the confidence level of the first biometric information is less than the third threshold. If the first biometric authentication is a doubtful determination, the continuation processing determination unit 52 decides to perform the second biometric authentication. The integrated determination unit 51 determines an OK determination if, for example, the first biometric authentication is a doubtful determination and the second biometric authentication is a confirmed determination.

[0062] The continuation processing determination unit 52 determines the authentication result of the first biometric authentication of the person being authenticated as a "verification judgment" if the score of the first biometric authentication is equal to or greater than the second threshold, and the confidence level of the first biometric information is less than the fourth threshold, and does not determine whether or not the person being authenticated is the person in question. If the first biometric authentication is a verification judgment, the continuation processing determination unit 52 decides to perform the second biometric authentication. The integrated determination unit 51 determines an OK judgment, for example, if the first biometric authentication is a verification judgment and the second biometric authentication is either a verification judgment or a confirmation judgment.

[0063] <Embodiment 1> Next, with reference to Figure 5, the continuous biometric authentication process according to Embodiment 1 will be described. Figure 5 is a flowchart of the continuous biometric authentication process according to Embodiment 1. Each process shown in Figure 5 is executed by the processor 22.

[0064] The processor 22 activates the sensor 20 (step St10).

[0065] The system configuration unit 53 obtains system configuration information from at least one of the memory 23 or the user database 30 (step St11). Based on the system configuration information, the processor 22 obtains biometric information to be used for first biometric authentication from the sensor 20. For example, if the processor 22 performs biometric authentication using facial biometric information as the first biometric authentication based on the system configuration information, it obtains imaging data of the face of the person to be authenticated from the face camera 201.

[0066] The modal processing unit 60 (or modal processing unit 70) performs modal authentication based on the biometric information acquired from the sensor 20 (step St12). The modal processing unit 60 (or modal processing unit 70) transmits the score of the first biometric authentication and the confidence level of the first biometric information calculated in step St12 to the continuous processing determination unit 52.

[0067] The continuation processing determination unit 52 determines whether to continue biometric authentication based on the score of the first biometric authentication and the confidence level of the first biometric information obtained in the processing of step St12 (step St13). For example, the continuation processing determination unit 52 determines whether the first biometric authentication will result in a confirmed judgment, a rejected judgment, a questionable judgment, or a verified judgment based on the score of the first biometric authentication and the confidence level of the first biometric information obtained in the processing of step St12. Based on the determined judgment result, the continuation processing determination unit 52 determines whether to continue biometric authentication.

[0068] If the continuation processing determination unit 52 determines in step St13 that biometric authentication should be continued (step St13, YES), it sets the continuation process (step St14). Setting the continuation process involves setting which biometric information to use for the second biometric authentication. The continuation processing determination unit 52 may decide which biometric information to use for the second biometric authentication based on a predetermined order. Alternatively, the continuation processing determination unit 52 may decide which biometric information to use for the second biometric authentication based on the reliability of the first biometric information. For example, if the first biometric information is imaging data related to the face and has a low reliability (for example, below the fourth threshold in the example shown in Figure 4), the continuation processing determination unit 52 will decide that the second biometric authentication will be biometric authentication using sound acquisition data related to speech. The method by which the continuation processing determination unit 52 determines the biometric information to be used for the second biometric authentication based on the reliability of the first biometric information is described in detail in Figure 6.

[0069] After setting the continuation process in step St14, the continuation processing determination unit 52 causes the modal processing unit 60 (or modal processing unit 70) to perform modal authentication (e.g., second biometric authentication) again.

[0070] If the continuation processing determination unit 52 determines in step St13 that biometric authentication should not be continued (step St13, NO), it transmits the determination result to the integrated determination unit 51. Based on the authentication result obtained from the continuation processing determination unit 52, the integrated determination unit 51 determines whether the person being authenticated is the person in question (i.e., whether it is an NG judgment or an OK judgment) (step St15).

[0071] If the integrated determination unit 51 determines that the person to be authenticated is the person in question (step St15, YES), it executes an OK process (step St16). The OK process may, for example, involve the integrated determination unit 51 displaying a screen on the display 21 indicating that the person to be authenticated is the person in question, or having the drive unit 40 open a gate installed in the authentication device 10 or an external device (not shown). Note that the OK process is just one example and is not limited to these.

[0072] If the integrated determination unit 51 determines that the person to be authenticated is not the person in question (step St15, NO), it executes an NG process (step St17). An NG process is, for example, the integrated determination unit 51 to display a screen on the display 21 indicating that the person to be authenticated is not the person in question, or to have the drive unit 40 close a gate installed in the authentication device 10 or an external device (not shown). Note that the examples of NG processes are just examples and are not limited to these.

[0073] This enables the authentication device 10 to perform more accurate authentication by determining whether or not to continue authentication considering the quality of the biometric information used for identity verification.

[0074] Next, with reference to Figure 6, we will explain the case in which a second biometric authentication is selected based on the first biometric authentication. Figure 6 is a conceptual diagram of the case in which a second biometric authentication is selected based on the first biometric authentication. Each biometric authentication in Figure 6 will be determined as confirmed, rejected, questionable, or verified according to the first, second, third, and fourth thresholds explained in Figure 4. The value of the first threshold is 30, the value of the second threshold is 70, the value of the third threshold is 30, and the value of the fourth threshold is 70. Note that the values ​​of each threshold are examples and are not limited to these. The biometric authentications in Figures 7, 8, and 10 will also be determined by the same thresholds.

[0075] Case CC illustrates an example of performing biometric authentication in a predetermined order.

[0076] Authentication Cer4 acquires biometric information related to the face of the person being authenticated and displays the result of biometric authentication by comparing it with the person's biometric information stored in the user DB30 beforehand. The score of Authentication Cer4 is 50, which is above the first threshold and below the second threshold. In Authentication Cer4, the biometric information related to the face of the person being authenticated was acquired in a dark environment where the face was unclear, so the confidence level is 50, which is above the third threshold and below the fourth threshold. Authentication Cer4 is judged as questionable, and the continuation processing determination unit 52 performs Authentication Cer5 as the second biometric authentication.

[0077] Authentication Cer5 acquires biometric information related to the fingerprint of the person to be authenticated and presents the result of biometric authentication by comparing it with the person's biometric information stored in the user DB30 beforehand. The score of Authentication Cer5 is 60, which is above the first threshold and below the second threshold. In Authentication Cer5, the biometric information related to the fingerprint was acquired in a dark environment where the fingerprint of the person to be authenticated was unclear, so the confidence level is 50, which is above the third threshold and below the fourth threshold. Authentication Cer5 is judged as questionable, and the continuation processing determination unit 52 performs Authentication Cer6 as a third biometric authentication (referred to as the third biometric authentication).

[0078] Authentication Cer6 acquires biometric information related to the speech of the person being authenticated and displays the result of biometric authentication by comparing it with the person's biometric information stored in the user DB30 beforehand. The score of Authentication Cer6 is 90, which is above the second threshold. In Authentication Cer6, the biometric information related to the speech was acquired in a quiet environment in which the speech of the person being authenticated could be clearly picked up, so the confidence level is 100, which is above the fourth threshold. Authentication Cer6 is determined to be confirmed, and the continuation processing determination unit 52 terminates biometric authentication and transmits the results of Authentication Cer4, Authentication Cer5, and Authentication Cer6 to the integrated determination unit 51.

[0079] The integrated determination unit 51 determines that the authentication of the person to be authenticated is OK based on the information obtained from the continuous processing determination unit 52.

[0080] Case CD shows an example of determining the second biometric authentication performed by the continuation processing determination unit 52 based on the first biometric authentication.

[0081] Authentication Cer7 acquires biometric information related to the face of the person being authenticated and displays the result of biometric authentication by comparing it with the person's biometric information stored in the user DB30 beforehand. The score of Authentication Cer7 is 50, which is above the first threshold and below the second threshold. In Authentication Cer7, the biometric information related to the face of the person being authenticated was acquired in a dark environment where the face was unclear, so the confidence level is 50, which is above the third threshold and below the fourth threshold. Authentication Cer7 is judged as questionable, and the continuation processing determination unit 52 performs second biometric authentication.

[0082] The continuation processing determination unit 52 determines that if the first biometric authentication is either a doubt determination or a verification determination, and the first biometric information is biometric information related to image data (e.g., face or fingerprint), the second biometric authentication will use biometric information related to the sound acquisition data of the speech. If the reliability of the biometric information related to the face, i.e., the image data, is low, it is expected that the reliability of the biometric information related to the fingerprint, which also uses image data, will also be low. Therefore, it is expected that using sound acquisition data for the second biometric authentication will result in higher reliability, and it may be possible to complete biometric authentication in fewer attempts than in case CC.

[0083] Authentication Cer8 acquires biometric information related to the speech of the person being authenticated and displays the authentication result after comparing it with the person's biometric information previously stored in the user DB30. The score of Authentication Cer8 is 90, which is above the second threshold. In Authentication Cer8, the biometric information related to the speech was acquired in a quiet environment in which the speech of the person being authenticated could be clearly picked up, so the confidence level is 100, which is above the fourth threshold. Authentication Cer8 is determined to be confirmed, and the continuation processing determination unit 52 terminates biometric authentication and transmits the results of Authentication Cer7 and Authentication Cer8 to the integrated determination unit 51.

[0084] The integrated determination unit 51 determines that the authentication of the person to be authenticated is OK based on the results of authentication Cer7 and authentication Cer8 obtained from the continuous processing determination unit 52.

[0085] As a result, the authentication device 10 according to Embodiment 1 can reduce the number of authentication steps and alleviate the complexity of user authentication by determining the second biometric information based on the result of the first biometric authentication.

[0086] Next, with reference to Figure 7, an example of a method for determining when the reliability of all biometric authentications is low will be explained. Figure 7 is a diagram showing an example of an authentication method when the reliability of all biometric authentications is low.

[0087] Case CE illustrates an example of a method for determining when the reliability of all biometric authentication methods is low.

[0088] Authentication Cer9 acquires biometric information related to the face of the person being authenticated and shows the result of biometric authentication by comparing it with the person's biometric information stored in the user DB30 beforehand. The score for Authentication Cer9 is 50, which is above the first threshold and below the second threshold. Authentication Cer9 is performed in an environment where the face of the person being authenticated is so dark that it is unclear, and the speech of the person being authenticated is so noisy that it is unclear. Therefore, the confidence level of Authentication Cer9 is 50, which is above the third threshold and below the fourth threshold. Authentication Cer4 is judged as questionable.

[0089] Authentication Cer10 acquires biometric information related to the fingerprint of the person being authenticated and shows the result of biometric authentication by comparing it with the person's biometric information stored in the user DB30 beforehand. The score of Authentication Cer10 is 60, which is above the first threshold and below the second threshold. Authentication Cer10 is performed in an environment that is so dark that the person being authenticated's face is unclear, and so noisy that the person being authenticated's speech is unclear. Therefore, the confidence level of Authentication Cer10 is 60, which is above the third threshold and below the fourth threshold. Authentication Cer10 is judged as questionable.

[0090] Authentication Cer11 acquires biometric information related to the speech of the person being authenticated and shows the result of biometric authentication by comparing it with the person's biometric information stored in the user DB30 beforehand. The score of Authentication Cer11 is 30, which is above the first threshold and below the second threshold. Authentication Cer11 is performed in an environment where the person being authenticated is too dark to see clearly and the person's speech is too noisy to hear clearly. Therefore, the confidence level of Authentication Cer11 is 30, which is above the third threshold and below the fourth threshold. Authentication Cer11 is judged as questionable.

[0091] The integrated determination unit 51 obtains the scores and confidence levels of authentication Cer9, authentication Cer10, and authentication Cer11 from the continuous processing determination unit 52. The integrated determination unit 51 sums the scores of authentication Cer9, authentication Cer10, and authentication Cer11, compares the summed score to a threshold, and determines OK if it is equal to or greater than the threshold. In addition, the integrated determination unit 51 may also consider the summed confidence levels of authentication Cer9, authentication Cer10, and authentication Cer11 and compare the summed confidence levels to a threshold when determining OK or NG.

[0092] As a result, the authentication device 10 can perform authentication with higher accuracy by making an integrated determination using multiple biometric authentication methods.

[0093] Next, referring to Figure 8, we will explain an example in which the person being authenticated is determined not to be the person in question based on the results of multiple biometric authentications. Figure 8 is a diagram illustrating an example in which the person being authenticated is determined not to be the person in question based on the results of multiple biometric authentications.

[0094] Case CF illustrates an example of a failure based on the results of multiple biometric authentication attempts.

[0095] Authentication Cer13 acquires biometric information related to the face of the person being authenticated and shows the result of biometric authentication by comparing it with the person's biometric information stored in the user DB30 beforehand. The score for Authentication Cer13 is 50, which is above the first threshold and below the second threshold. In Authentication Cer13, the biometric information related to the face of the person being authenticated was acquired in a dark environment where the face of the person being authenticated was unclear, so the confidence level is 50, which is above the third threshold and below the fourth threshold. Authentication Cer13 is judged as questionable.

[0096] Authentication Cer14 acquires biometric information related to the fingerprint of the person being authenticated and shows the result of biometric authentication by comparing it with the person's biometric information stored in the user DB30 beforehand. The score for Authentication Cer14 is 60, which is above the first threshold and below the second threshold. In Authentication Cer14, the biometric information related to the fingerprint was acquired in a dark environment where the fingerprint of the person being authenticated was unclear, so the confidence level is 60, which is above the third threshold and below the fourth threshold. Authentication Cer14 is judged as questionable.

[0097] Authentication Cer15 acquires biometric information related to the speech of the person being authenticated and compares it with the person's biometric information stored in the user DB30 beforehand to show the result of biometric authentication. The score for Authentication Cer15 is 20, which is below the first threshold. In Authentication Cer15, the biometric information related to the speech was acquired in a quiet environment where the speech of the person being authenticated could be clearly picked up, so the confidence level is 100, which is above the fourth threshold. Authentication Cer15 is rejected.

[0098] The integrated determination unit 51 obtains the results of authentication Cer13, authentication Cer14, and authentication Cer15 from the continuation processing determination unit 52. If at least one of the multiple biometric authentications has a high confidence level but a low score (for example, a rejection), the integrated determination unit 51 determines that the identity verification authentication is NG (Not Good).

[0099] <Embodiment 2> In Embodiment 2, the authentication device 10 acquires information about the external environment before starting the first biometric authentication, and determines the biometric information to be used for the first biometric authentication based on the acquired external environment.

[0100] Figure 9 illustrates the continuous biometric authentication process according to Embodiment 2. Figure 9 is a flowchart of the continuous biometric authentication process according to Embodiment 2. Each process in Figure 9 is executed by the processor 22. For processes in Figure 9 that are the same as those in Figure 5, the same reference numerals are used as in Figure 5 and their explanations are omitted.

[0101] After activating the sensor 20 in step St10, the processor 22 instructs the sensor 20 to acquire information about the surrounding environment of the authentication device 10. This information about the surrounding environment includes, for example, brightness or noise level. Based on the signal from the processor 22, the sensor 20 acquires the information about the surrounding environment and outputs the acquired information to the processor 22. The processor 22 acquires the information about the surrounding environment (step St20).

[0102] The continuation processing determination unit determines the order of biometric authentication based on the surrounding environment information acquired in step St20 and sends it to the system setting unit 53. The system setting unit 53 sets the order of biometric authentication (step St21). The processing from step St21 onwards in Figure 9 is the same as the processing in the flowchart in Figure 5.

[0103] Next, with reference to Figure 10, we will explain an example of determining the first biometric authentication based on information about the surrounding environment acquired before the authentication process begins. Figure 10 is a diagram illustrating an example of determining the first biometric authentication based on information about the surrounding environment acquired before the authentication process begins.

[0104] Case CC shown in Figure 10 is the same as Case CC in Figure 6, so its explanation is omitted.

[0105] Case CH demonstrates an example where an OK result is obtained using the first biometric authentication method, which is determined based on information about the surrounding environment acquired before the authentication process begins.

[0106] The continuation processing determination unit 52 acquires information about the surrounding environment before starting the first biometric authentication. If the surrounding environment is dark, it performs the first biometric authentication using speech instead of facial and fingerprint biometric authentication. However, if the surrounding environment is noisy, the continuation processing determination unit 52 performs facial and fingerprint biometric authentication as the first biometric authentication.

[0107] Authentication Cer16 acquires biometric information related to the speech of the person being authenticated and displays the result of biometric authentication by comparing it with the person's biometric information stored in the user DB30 beforehand. The score of Authentication Cer16 is 90, which is above the second threshold. In Authentication Cer16, the biometric information related to the speech was acquired in a quiet environment in which the speech of the person being authenticated could be clearly picked up, so the confidence level is 100, which is above the fourth threshold. Authentication Cer16 is determined to be confirmed, and the continuation processing determination unit 52 terminates biometric authentication and transmits the result of Authentication Cer16 to the integrated determination unit 51.

[0108] The integrated determination unit 51 determines that the authentication of the person to be authenticated is OK based on the result of the authentication Cer16 obtained from the continuous processing determination unit 52.

[0109] As a result, the authentication device 10 determines the first biometric authentication according to the surrounding environment, thereby reducing the number of authentication steps and alleviating the complexity of authenticating the person being authenticated.

[0110] Next, we will explain an example of the reliability of each type of biological information by referring to Figure 11. Figure 11 is a diagram showing an example of the reliability of each type of biological information.

[0111] For example, a "high" confidence level represents a level above the fourth threshold in Figure 4, a "medium" level represents a level above the third threshold but below the fourth threshold, and a "low" level represents a level below the third threshold. Note that the possible ranges for "high," "medium," and "low" confidence levels are just examples and are not limited to those described above.

[0112] The factors used to evaluate facial reliability (hereinafter referred to as "reliability evaluation factors") include, for example, facial orientation, lighting conditions, facial expression, partial concealment, or registration quality. However, these are not the only examples of facial reliability evaluation factors.

[0113] When "face orientation" is a factor in evaluating facial reliability, the reliability changes depending on the angle of the person's face relative to the face camera 201. If the face is facing directly towards the face camera 201, the reliability is "high". If the face is at a moderate angle relative to the face camera 201, the reliability is "medium". If the face is at a large angle relative to the face camera 201, the reliability is "low".

[0114] When the facial reliability evaluation factor is "lighting conditions," the reliability changes depending on at least one of the following: the direction of the light illuminating the subject's face or the illuminance. If the light is illuminating the subject's face from the front with appropriate illuminance, the reliability will be "high." If the light is illuminating the subject's face from anywhere other than the front, the reliability will be "medium." If the light is illuminating the subject's face with low illuminance, the reliability will be "low."

[0115] If the facial reliability evaluation factor is "facial expression," the reliability changes depending on the facial expression of the person being authenticated. If the person being authenticated has a neutral expression, the reliability will be "high." If the person being authenticated has a smile or other facial expression, the reliability will be "medium."

[0116] When the facial reliability evaluation factor is "partial concealment," the reliability changes depending on the percentage of the subject's face that is concealed by a mask, sunglasses, hat, etc. If the subject's face is not concealed, the reliability is "high." If a small portion of the subject's face is concealed, the reliability is "medium." If a large portion of the subject's face is concealed, the reliability is "low."

[0117] When the facial reliability evaluation factor is "Registered Quality," the reliability changes depending on whether the person being authenticated has siblings. If the person being authenticated has siblings, the reliability is "Medium." If the person being authenticated is a twin, the reliability is "Low."

[0118] Next, we will explain the factors for evaluating the reliability of speech. These factors include, for example, the surrounding environment, utterance length, complexity, or recording quality. However, these are not the only examples of factors that can be used to evaluate the reliability of speech.

[0119] If the reliability evaluation factor for voice is "surrounding environment," the reliability will change depending on at least one of the ambient noise surrounding the authentication device 10 or its installation location. If the surrounding environment is quiet, the reliability will be "high." If the surrounding environment has ambient noise and the authentication device 10 is installed outdoors, the reliability will be "medium." If the surrounding environment has ambient noise and the authentication device 10 is installed indoors, the reliability will be "low."

[0120] When the reliability evaluation factor for audio is "utterance length," the reliability changes depending on the length of the utterance of the person being authenticated (hereinafter referred to as "utterance length") contained in the recorded audio data. If the recorded audio data has a sufficient utterance length, the reliability will be "high." If the recorded audio data has a moderate utterance length, the reliability will be "medium." If the recorded audio data has only a short utterance length, the reliability will be "low."

[0121] When the confidence rating factor for audio is "complexity," the confidence level changes depending on the number or pattern of consonants and vowels included in the recorded data. If the recorded data contains a sufficient number of consonants and vowels, the confidence level will be "high." If the recorded data contains a limited number of consonant and vowel patterns, the confidence level will be "medium."

[0122] When the reliability evaluation factor for audio is "registration quality," the reliability changes depending on the quality of the recorded audio data. If the registration quality of the recorded audio data is low and the speech length is short, the reliability will be "medium."

[0123] Next, we will explain the factors for evaluating the reliability of fingerprints. These factors include, for example, the acquisition area or image quality. However, these are not the only examples of factors that can be used to evaluate the reliability of fingerprints.

[0124] If the reliability evaluation factor for fingerprints is "acquisition area," the reliability changes depending on the area of ​​the acquired fingerprint relative to the size of the person being authenticated's finger. If almost the entire finger's fingerprint is acquired, the reliability is "high." If part of the fingerprint is missing, the reliability is "medium." If only a portion of the fingerprint is acquired, the reliability is "low."

[0125] When the reliability evaluation factor for fingerprints is "image quality," the reliability changes depending on the contrast or focus. If the contrast is good and the focus is correct, the reliability will be "high." If the contrast is insufficient and the image is out of focus, resulting in blurry image data, the reliability will be "low."

[0126] As described above, the authentication device 10 according to this embodiment includes an acquisition unit (e.g., sensor 20) that acquires at least the first biometric information of the object to be authenticated used for first biometric authentication and the first surrounding environment information at the time of acquisition of the first biometric information. The authentication device 10 includes a score calculation unit (e.g., score calculation unit 62 or score calculation unit 72) that calculates a score for first biometric authentication using the first biometric information. The authentication device 10 includes a reliability determination unit (e.g., reliability determination unit 61 or reliability determination unit 71) that determines the reliability of the first biometric information from the first surrounding environment information. The authentication device 10 includes a continuation processing determination unit (e.g., continuation processing determination unit 52) ​​that determines whether or not to perform second biometric authentication using second biometric information different from the first biometric information, based on the score of first biometric authentication and the reliability of the first biometric information.

[0127] As a result, the authentication device 10 according to this embodiment can determine whether or not to perform second biometric authentication based on the score of the first biometric authentication and the reliability of the first biometric information, thereby achieving highly accurate authentication that takes into account the quality of the biometric information used for identity verification.

[0128] Furthermore, in this embodiment, if the score of the first biometric authentication is less than the first threshold and the confidence level of the first biometric information is at or above the third threshold, the authentication result of the authentication target is determined to be the first determination that the authentication target is not the person in question, and the authentication of the authentication target is terminated without performing the second biometric authentication. As a result, the authentication device 10 can terminate the authentication of the person being authenticated without performing the second biometric authentication if the confidence level of the first biometric authentication is significantly low, thereby reducing the complexity of the authentication of the person being authenticated.

[0129] Furthermore, in this embodiment, if the score of the first biometric authentication is equal to or greater than the second threshold and the confidence level of the first biometric information is equal to or greater than the fourth threshold, the authentication result of the authentication target is determined to be the real person (second determination), and the authentication of the authentication target is terminated without performing the second biometric authentication. As a result, the authentication device 10 can terminate the authentication of the person being authenticated without performing the second biometric authentication if the confidence level of the first biometric authentication is sufficiently high, thereby reducing the complexity of the authentication of the person being authenticated.

[0130] Furthermore, the continuation processing determination unit in this embodiment determines that if the score of the first biometric authentication is equal to or greater than the first threshold but less than the second threshold, or if the score of the first biometric authentication is less than the first threshold and the confidence level of the first biometric information is less than the third threshold, the authentication result of the authentication target is determined to be the third determination, and the second biometric authentication is to be performed further. As a result, the authentication device 10 can perform the second biometric authentication if the score of the first biometric authentication is not sufficiently high, thereby achieving highly accurate authentication that takes into account the quality of the biometric information used for identity verification.

[0131] Furthermore, if the continuous processing determination unit in this embodiment determines the authentication result to be a third determination, it further performs a second biometric authentication using biometric information that is expected to have a higher reliability. As a result, the authentication device 10 can select a biometric authentication that is expected to have a higher reliability and perform the second biometric authentication, thereby reducing unnecessary authentication and alleviating the complexity of authenticating the person being authenticated.

[0132] Furthermore, the continuation processing determination unit in this embodiment determines that if the score of the first biometric authentication is equal to or greater than the second threshold, and the confidence level of the first biometric information is less than the fourth threshold, the authentication result of the authentication target is determined to be the fourth determination, and the second biometric authentication is to be performed further. As a result, the authentication device 10 can perform the second biometric authentication even if the score of the first biometric authentication is sufficiently high but the confidence level is low, thereby achieving highly accurate authentication that takes into account the quality of the biometric information used for identity verification.

[0133] Furthermore, the continuous processing determination unit in this embodiment determines the authentication result of the authentication target as a third determination if the score of the first biometric authentication is equal to or greater than the first threshold but less than the second threshold, or if the score of the first biometric authentication is less than the first threshold and the confidence level of the first biometric information is less than the third threshold. The continuous processing determination unit determines the authentication result of the authentication target as a fourth determination if the score of the first biometric authentication is equal to or greater than the second threshold and the confidence level of the first biometric information is less than the fourth threshold. In the case of the third and fourth determinations, the continuous processing determination unit determines to perform the second biometric authentication further. As a result, the authentication device 10 can perform the second biometric authentication if the score of the first biometric authentication is not sufficiently high or if the score of the first biometric authentication is sufficiently high but the confidence level is low, thereby achieving highly accurate authentication that takes into account the quality of the biometric information used for identity verification.

[0134] Furthermore, the authentication device 10 according to this embodiment further includes an integrated determination unit (for example, an integrated determination unit 51) that determines that the person being authenticated is the real person if both the authentication result of the first biometric authentication using the first biometric information and the authentication result of the second biometric authentication using the second biometric information are of the third or fourth determination, and the sum of the scores of the first and second biometric authentications is equal to or greater than a predetermined value. As a result, the authentication device 10 can make an integrated determination that the person being authenticated is the real person even if the scores of the first and second biometric authentications are not sufficiently high, as long as the sum of the scores is equal to or greater than a predetermined value. This enables the authentication device 10 to achieve highly accurate authentication using multiple biometric information.

[0135] Furthermore, in this embodiment, the integrated determination unit determines that the person being authenticated is not the person in question if both the authentication result of the first biometric authentication using the first biometric information and the authentication result of the second biometric authentication using the second biometric information are of the third or fourth determination, and the sum of the scores is less than a predetermined value. As a result, the authentication device 10 can make an integrated determination that the person being authenticated is not the person in question even if the scores of the first and second biometric authentication are not sufficiently high, as long as the sum of the scores is less than a predetermined value. This enables the authentication device 10 to achieve highly accurate authentication using multiple biometric information.

[0136] Furthermore, the continuation processing determination unit according to this embodiment determines the execution order of the first biometric authentication and the second biometric authentication based on the first surrounding environment information. As a result, the authentication device 10 can determine the execution order of the first biometric authentication and the second biometric authentication from the first surrounding environment information, and can prioritize the execution of authentication that is suitable for the first surrounding environment information.

[0137] Furthermore, the acquisition unit in this embodiment acquires the second biometric information of the authentication target to be used for the second biometric authentication and the second surrounding environment information at the time of acquisition of the second biometric information. Based on the first and second surrounding environment information acquired in advance by the acquisition unit, the continuation processing determination unit starts with the first or second biometric authentication corresponding to the higher of the reliability levels of the first biometric information and the second biometric information. As a result, the authentication device 10 can prioritize biometric authentication that is expected to have a high reliability level based on the first surrounding environment information, thereby reducing the number of authentications and alleviating the complexity of authenticating the person to be authenticated.

[0138] Furthermore, the first biometric information in this embodiment is image data of the face to be authenticated, fingerprint data, or spoken voice data. This allows the authentication device 10 to perform biometric authentication using each of these biometric information.

[0139] Furthermore, the integrated determination unit according to this embodiment calculates the false acceptance rate of the authentication target based on the authentication result of the first biometric authentication and the authentication result of the second biometric authentication. As a result, the authentication device 10 can present the calculated false acceptance rate to the authentication target and inform them of the accuracy of the authentication result.

[0140] Furthermore, the reliability determination unit according to this embodiment calculates the reliability of the first biometric information based on the orientation of the face to be authenticated, lighting conditions, facial expression, length of speech, number of phonemes, image quality, or fingerprint acquisition area. As a result, the authentication device 10 can determine the reliability of the first biometric information based on various conditions.

[0141] While embodiments have been described above with reference to the attached drawings, this disclosure is not limited to such examples. It is clear to those skilled in the art that various modifications, alterations, substitutions, additions, deletions, and equivalents can be conceived within the scope of the claims, and these are also understood to fall within the technical scope of this disclosure. Furthermore, the components of the embodiments described above can be combined in any way without departing from the spirit of the invention. [Industrial applicability]

[0142] The technology disclosed herein is useful as an authentication device and authentication method that achieves high-precision authentication while taking into account the quality of biometric information used for identity verification. [Explanation of symbols]

[0143] 1. Authentication System 10 Authentication device 20 sensors 21 displays 22 processors 23 memory 24 Input devices 25 Communication IF 30 User B 40 Drive unit 50 Overall control unit 51 Integrated determination unit 52 Continuation Processing Determination Unit 53 System Settings Section 54 User characteristics determination section 60,70 Modal Processing Unit 61,71 Confidence Determination Unit 62,72 Score calculation section 201 Face Camera 202 Fingerprint Camera 203 Mike CA,CB,CC,CD,CE,CF,CH Case Cer1, Cer2, Cer3, Cer4, Cer5, Cer6, Cer7, Cer8, Cer9, Cer10, Cer11, Cer13, Cer14, Cer15, Cer16 Authentication

Claims

1. An acquisition unit that acquires at least the first biometric information of the target to be authenticated for use in first biometric authentication and the first surrounding environment information at the time of acquisition of the first biometric information, A score calculation unit that calculates the score of the first biometric authentication using the first biometric information, A reliability determination unit that determines the reliability of the first biological information from the first surrounding environment information, The system includes a continuation processing determination unit that determines whether or not to perform a second biometric authentication using second biometric information different from the first biometric information, based on the score of the first biometric authentication and the confidence level of the first biometric information. The continuation processing determination unit determines that if the score of the first biometric authentication is greater than or equal to the first threshold but less than the second threshold, or if the score of the first biometric authentication is less than the first threshold and the confidence level of the first biometric information is less than the third threshold, the authentication result of the authentication target is determined to be the third determination, and the second biometric authentication is performed further. Authentication device.

2. If the continuation processing determination unit determines the authentication result to be the third determination, it further performs the second biometric authentication using biometric information that is expected to have a high reliability. The authentication device according to claim 1.

3. An acquisition unit that acquires at least the first biometric information of the target to be authenticated for use in first biometric authentication and the first surrounding environment information at the time of acquisition of the first biometric information, A score calculation unit that calculates the score of the first biometric authentication using the first biometric information, A reliability determination unit that determines the reliability of the first biological information from the first surrounding environment information, The system includes a continuation processing determination unit that determines whether or not to perform a second biometric authentication using second biometric information different from the first biometric information, based on the score of the first biometric authentication and the confidence level of the first biometric information. The continuation processing determination unit determines, if the score of the first biometric authentication is equal to or greater than the second threshold, and the confidence level of the first biometric information is less than the fourth threshold, to set the authentication result of the authentication target as the fourth determination and to perform the second biometric authentication again. If the score of the second biometric authentication is equal to or greater than the second threshold, it is determined that the person being authenticated is the person in question. Authentication device.

4. An acquisition unit that acquires at least the first biometric information of the target to be authenticated for use in first biometric authentication and the first surrounding environment information at the time of acquisition of the first biometric information, A score calculation unit that calculates the score of the first biometric authentication using the first biometric information, A reliability determination unit that determines the reliability of the first biological information from the first surrounding environment information, The system includes a continuation processing determination unit that determines whether or not to perform a second biometric authentication using second biometric information different from the first biometric information, based on the score of the first biometric authentication and the confidence level of the first biometric information. The continuation processing determination unit, If the score of the first biometric authentication is greater than or equal to the first threshold but less than the second threshold, or if the score of the first biometric authentication is less than the first threshold and the confidence level of the first biometric information is less than the third threshold, the authentication result of the authentication target shall be determined as the third determination. If the score of the first biometric authentication is equal to or greater than the second threshold, and the confidence level of the first biometric information is less than the fourth threshold, the authentication result of the authentication target is determined to be the fourth determination. In the case of the third and fourth determinations, it is determined that the second biometric authentication will be performed further. Authentication device.

5. The system further includes an integrated determination unit that determines that the person being authenticated is the person in question if both the authentication result of the first biometric authentication using the first biometric information and the authentication result of the second biometric authentication using the second biometric information are either the third determination or the fourth determination, and the sum of the scores of the first and second biometric authentications is equal to or greater than a predetermined value. The authentication device according to claim 4.

6. The integrated determination unit determines that the person being authenticated is not the person in question if both the authentication result of the first biometric authentication using the first biometric information and the authentication result of the second biometric authentication using the second biometric information are either the third determination or the fourth determination, and the sum of the respective scores is less than a predetermined value. The authentication device according to claim 5.

7. The integrated determination unit calculates the false acceptance rate of the authentication target based on the authentication result of the first biometric authentication and the authentication result of the second biometric authentication. The authentication device according to claim 5.

8. An authentication method performed by an authentication device that authenticates the identity of the person being authenticated, The system acquires at least the first biometric information of the subject to be authenticated, which is used for the first biometric authentication, and the first surrounding environment information at the time the first biometric information was acquired. The score of the first biometric authentication using the first biometric information is calculated. The reliability of the first biological information is determined from the first surrounding environment information. Based on the score of the first biometric authentication and the confidence level of the first biometric information, a determination is made as to whether or not to perform a second biometric authentication using second biometric information different from the first biometric information. If the score of the first biometric authentication is greater than or equal to the first threshold but less than the second threshold, or if the score of the first biometric authentication is less than the first threshold and the confidence level of the first biometric information is less than the third threshold, the authentication result of the authentication target is determined to be the third determination, and it is determined to perform the second biometric authentication further. Authentication method.

9. An authentication method performed by an authentication device that authenticates the identity of the person being authenticated, The system acquires at least the first biometric information of the subject to be authenticated, which is used for the first biometric authentication, and the first surrounding environment information at the time the first biometric information was acquired. The score of the first biometric authentication using the first biometric information is calculated. The reliability of the first biological information is determined from the first surrounding environment information. Based on the score of the first biometric authentication and the confidence level of the first biometric information, a determination is made as to whether or not to perform a second biometric authentication using second biometric information different from the first biometric information. If the score of the first biometric authentication is equal to or greater than the second threshold, and the confidence level of the first biometric information is less than the fourth threshold, the authentication result of the authentication target is determined to be the fourth determination, and it is determined to perform the second biometric authentication again. If the score of the second biometric authentication is equal to or greater than the second threshold, it is determined that the person being authenticated is the person in question. Authentication method.

10. An authentication method performed by an authentication device that authenticates the identity of the person being authenticated, The system acquires at least the first biometric information of the subject to be authenticated, which is used for the first biometric authentication, and the first surrounding environment information at the time the first biometric information was acquired. The score of the first biometric authentication using the first biometric information is calculated. The reliability of the first biological information is determined from the first surrounding environment information. Based on the score of the first biometric authentication and the confidence level of the first biometric information, a determination is made as to whether or not to perform a second biometric authentication using second biometric information different from the first biometric information. If the score of the first biometric authentication is greater than or equal to the first threshold but less than the second threshold, or if the score of the first biometric authentication is less than the first threshold and the confidence level of the first biometric information is less than the third threshold, the authentication result of the authentication target shall be determined as the third determination. If the score of the first biometric authentication is equal to or greater than the second threshold, and the confidence level of the first biometric information is less than the fourth threshold, the authentication result of the authentication target is determined to be the fourth determination. In the case of the third and fourth determinations, it is determined that the second biometric authentication will be performed further. Authentication method.

Images