Registration information generating device, matching information generating device, matching system, registration information generating method, matching information generating method, similarity calculation method, registration information generating program, and matching information generating program

Abstract

[Problem] To realize cancelable conversion that can both suppress the deterioration of verification accuracy and reduce the risk of information leakage. [Solution] A registration information generation device (1D) that generates registration information on the basis of registration-side input information, wherein the registration-side input information includes elements (S) corresponding to a first set and / or elements (S(hat)) corresponding to a second set, intermediate information is generated by performing a distribution process of secretly distributing elements corresponding to a first set of inputted information and a replication process of replicating elements corresponding to a second set of inputted information, and registration information is generated by replacing the order of elements included in the intermediate information using a replacement key.

Description

[Technical field]

[0001] The present invention relates to a registration information generating device, a matching information generating device, a matching system, a registration information generating method, a matching information generating method, a similarity calculation method, a registration information generating program, and a matching information generating program. [Background technology]

[0002] In recent years, biometric authentication technology has been widely used as one of the user authentication methods, which performs authentication by matching biometric information such as fingerprints, veins, irises, and facial images. In the biometric authentication technology, feature amounts are extracted from the biometric information of the user and stored as registered information, and the feature amounts of the biometric information of the user input during authentication are matched against the registered information to determine whether or not the user can be authenticated.

[0003] Compared with authentication technologies using IC cards, etc., biometric authentication technology has the advantage of being less prone to loss or theft, but even if biometric information is leaked, it cannot be discarded or updated. In other words, the leakage of biometric information not only poses a problem regarding personal information of the living body related to the leaked biometric information, but also poses a problem of compromising the security of authentication systems that use the biometric information of the living body related to the leaked biometric information.

[0004] To address these issues, a technology called "cancellable biometrics" has been proposed that makes it possible to invalidate registered information in order to protect the biometric information of users.

[0005] In cancelable biometrics, the elements of the feature extracted from the biometric information are encrypted by rearranging them using a secret key to generate registration information. In addition, the elements of the feature extracted from the biometric information input during authentication are also encrypted using the same secret key as that used during registration to generate matching information. In other words, when recreating the registration information, it is possible to invalidate the already generated registration information by using a different secret key. The encryption in cancelable biometrics is called "cancelable transformation" due to the nature of such registration information. As an example of cancelable transformation, Non-Patent Document 1 proposes a method of dividing an image of feature points obtained from a fingerprint or the like into blocks and rearranging the divided images using a secret key to generate registration information and matching information. [Prior art documents] [Non-patent literature]

[0006] [Non-Patent Document 1] NKRatha, et al., “Enhancing security and privacy in biometrics-based authentication systems”, IBM Systems Journal 40(3), 614-634, 2001 Summary of the Invention [Problem to be solved by the invention]

[0007] As described above, in the cancelable transformation, the elements of the feature extracted from the biometric information are transformed using the same secret key at the time of enrollment and at the time of matching. Therefore, in order to reduce the risk of leakage of the biometric information, it is desirable to match the enrollment information with the matching information while keeping the information in the cancelable transformed state. In addition, when matching the enrollment information with the matching information, it is desirable to perform the cancelable transformation so that matching accuracy equivalent to that of conventional biometric authentication technology can be obtained.

[0008] The object of the present invention has been made in consideration of the above-mentioned problems, and is to provide a registration information generating device, a matching information generating device, a matching system, a registration information generating method, a matching information generating method, a similarity calculation method, a registration information generating program, and a matching information generating program that realizes a cancellable conversion that can simultaneously suppress deterioration of matching accuracy and reduce the risk of information leakage. [Means for solving the problem]

[0009] The registration information generation device of the present invention is a registration information generation device that generates registration information based on registration side input information input for registration, wherein the registration side input information includes at least one of an element corresponding to a first set and an element corresponding to a second set different from the first set, and the registration information generation device is equipped with a registration side intermediate information generation unit that generates registration side intermediate information from the registration side input information by performing a registration side distribution process that secretly distributes elements corresponding to the first set of the input information and a registration side duplication process that duplicates elements corresponding to the second set of the input information, and a registration side replacement processing unit that performs a registration side replacement process that replaces the order of elements included in the input information using a replacement key, and the registration information is generated by replacing the order of elements included in the registration side intermediate information using the replacement key.

[0010] The registration information generation device of the present invention is a registration information generation device that generates registration information based on registration side input information input for registration, and is equipped with a distribution processing unit that generates secret sharing information from the registration side input information by performing a distribution process that secretly distributes elements included in the input information, and a registration side replacement processing unit that performs a registration side replacement process that replaces the order of the elements included in the input information using a replacement key, and the registration information is generated by replacing the order of the elements included in the secret sharing information using the replacement key.

[0011] A matching information generation device of the present invention is a matching information generation device that generates matching information based on matching-side input information input for matching, wherein the matching-side input information includes at least one of elements corresponding to a first set and elements corresponding to a second set different from the first set, and the matching information generation device comprises a matching-side intermediate information generation unit that generates matching-side intermediate information from the matching-side input information by performing a matching-side duplication process that duplicates elements corresponding to the first set of the input information and a matching-side distribution process that secretly distributes elements corresponding to the second set of the input information, and a matching-side permutation processing unit that performs a matching-side permutation process that permutes the order of elements included in the input information using a substitution key, and the matching information is generated by permuting the order of elements included in the matching-side intermediate information using the substitution key.

[0012] The matching information generation device of the present invention is a matching information generation device that generates matching information based on matching-side input information input for matching, and includes a duplication processing unit that generates duplicate information from the matching-side input information by performing a duplication process that duplicates elements included in the input information, and a matching-side substitution processing unit that performs a matching-side substitution process that substitutes the order of the elements included in the input information using a substitution key, and the matching information is generated by substituting the order of the elements included in the duplicate information using the substitution key.

[0013] A matching system of the present invention includes a registration information generation device that generates registration information based on registration side input information input for registration, a matching information generation device that generates matching information based on matching side input information input for matching, and a matching device that matches the registration side input information with the matching side input information, wherein the registration side input information and the matching side input information each include at least one of an element corresponding to a first set and an element corresponding to a second set different from the first set, and the registration information generation device generates registration side intermediate information from the registration side input information by performing a registration side distribution process that secretly distributes elements corresponding to the first set of the input information and a registration side duplication process that duplicates elements corresponding to the second set of the input information, and a registration side replacement process that replaces the order of elements included in the input information using a replacement key. the matching information generation device comprises a matching-side intermediate information generation unit that generates matching-side intermediate information from the matching-side input information by performing a matching-side duplication process that duplicates elements corresponding to the first set of input information and a matching-side distribution process that secretly distributes elements corresponding to the second set of input information, and a matching-side permutation processing unit that performs a matching-side permutation process that permutes an order of elements included in the input information using the substitution key, and the matching device comprises a similarity calculation unit that calculates a similarity between the registration side input information and the matching side input information based on the registration information and the matching information, and the registration information is generated by permuting the order of elements included in the registration side intermediate information using the substitution key, and the matching information is generated by permuting the order of elements included in the matching side intermediate information using the substitution key.

[0014] The verification system of the present invention is a verification system comprising: a registration information generation device that generates registration information based on registration side input information input for registration; a matching information generation device that generates matching information based on matching side input information input for matching; and a verification device that matches the registration side input information with the matching side input information, wherein the registration information generation device comprises a distribution processing unit that generates shared information from the registration side input information by performing a distribution process that secretly distributes elements included in the input information; and a registration side replacement processing unit that performs a registration side replacement process that replaces the order of elements included in the input information using a replacement key, and the matching information generation device generates matching information based on the input information. and a matching-side permutation processing unit that performs a permutation process of permuting elements included in input information to generate duplicate information from the matching-side input information, and a matching-side permutation processing unit that performs a permutation process of permuting an order of elements included in input information using the substitution key, wherein the matching device comprises a similarity calculation unit that calculates a similarity between the registration side input information and the matching side input information based on the registration information and the matching information, wherein the registration information is generated by permuting an order of elements included in the shared information using the substitution key, and the matching information is generated by permuting an order of elements included in the duplicate information using the substitution key.

[0015] The registration information generation method of the present invention is a registration information generation method for generating registration information based on registration side input information input for registration, which includes at least one of elements corresponding to a first set and elements corresponding to a second set different from the first set, and includes the steps of generating registration side intermediate information from the registration side input information by performing a registration side distribution process that secretly distributes elements corresponding to the first set of the input information and a registration side duplication process that duplicates elements corresponding to the second set of the input information, and performing a registration side replacement process that replaces the order of elements included in the input information using a replacement key, and generates the registration information by replacing the order of elements included in the registration side intermediate information using the replacement key.

[0016] The registration information generation method of the present invention is a registration information generation method for generating registration information based on registration side input information input for registration, and includes the steps of generating secret sharing information from the registration side input information by performing a distribution process that secretly distributes elements contained in the input information, and performing a registration side replacement process that replaces the order of the elements contained in the input information using a replacement key, and generates the registration information by replacing the order of the elements contained in the secret sharing information using the replacement key.

[0017] A matching information generation method of the present invention is a method for generating matching information based on matching-side input information input for matching, which includes at least one of elements corresponding to a first set and elements corresponding to a second set different from the first set, and includes the steps of: generating matching-side intermediate information from the matching-side input information by performing a matching-side duplication process for duplication of elements corresponding to the first set of the input information and a matching-side distribution process for secretly sharing elements corresponding to the second set of the input information; and performing a matching-side substitution process for replacing the order of elements included in the input information using a substitution key, and generating the matching information by replacing the order of elements included in the matching-side intermediate information using the substitution key.

[0018] A matching information generation method of the present invention is a matching information generation method for generating matching information based on matching side input information input for matching, and includes the steps of: generating duplicated information from the matching side input information by performing a duplication process that duplicates elements included in the input information; and performing a matching side replacement process that replaces the order of the elements included in the input information using a replacement key, and generates the matching information by replacing the order of the elements included in the duplicated information using the replacement key.

[0019] A similarity calculation method of the present invention is a similarity calculation method in a matching system including a registration information generation device that generates registration information based on registration side input information input for registration, a matching information generation device that generates matching information based on matching side input information input for matching, and a matching device that matches the registration side input information with the matching side input information, wherein the registration side input information and the matching side input information each include at least one of elements corresponding to a first set and elements corresponding to a second set different from the first set, and the registration information generation device performs a registration side distribution process that secretly distributes elements corresponding to the first set of the input information and a registration side duplication process that duplicates elements corresponding to the second set of the input information, thereby generating registration side intermediate information from the registration side input information and rearranging the order of elements included in the input information and performing a registration side replacement process in which an order of elements included in the registration side intermediate information is replaced using the replacement key, generating the registration information by replacing an order of elements included in the registration side intermediate information using the replacement key, the matching information generation device performing a matching side duplication process in which elements corresponding to the first set of input information are duplicated and a matching side distribution process in which elements corresponding to the second set of input information are secretly distributed to generate matching side intermediate information from the matching side input information, and performing a matching side replacement process in which an order of elements included in the input information is replaced using the replacement key, generating the matching information by replacing an order of elements included in the matching side intermediate information using the replacement key, and the matching device calculating a similarity between the registration side input information and the matching side input information based on the registration information and the matching information.

[0020] A similarity calculation method of the present invention is a similarity calculation method in a verification system including a registration information generation device that generates registration information based on registration side input information input for registration, a verification information generation device that generates verification information based on verification side input information input for verification, and a verification device that verifies the registration side input information and the verification side input information, the registration information generation device generating shared information from the registration side input information by performing a sharing process for secretly sharing elements included in the input information, and performing a registration side permutation process for permuting the order of the elements included in the input information using a substitution key, the matching information generation device generates the registration information by replacing the order of elements included in the input information using the substitution key, and the matching information generation device performs a duplication process to duplicate elements included in the input information, thereby generating duplicate information from the matching side input information, and performs a matching side substitution process to replace the order of the elements included in the input information using the substitution key, the matching information is generated by replacing the order of the elements included in the duplicate information using the substitution key, and the matching device calculates a similarity between the registration side input information and the matching side input information based on the registration information and the matching information.

[0021] The registration information generation program of the present invention is a registration information generation program that generates registration information based on registration side input information input for registration, which includes at least one of elements corresponding to a first set and elements corresponding to a second set different from the first set, and causes a computer to execute the following steps: generating registration side intermediate information from the registration side input information by performing a registration side distribution process that secretly distributes elements corresponding to the first set of the input information and a registration side duplication process that duplicates elements corresponding to the second set of the input information; and performing a registration side replacement process that replaces the order of elements included in the input information using a replacement key, and generates the registration information by replacing the order of elements included in the registration side intermediate information using the replacement key.

[0022] The registration information generation program of the present invention is a registration information generation program that generates registration information based on registration side input information input for registration, and causes a computer to execute the following steps: generating secret sharing information from the registration side input information by performing a distribution process that secretly distributes elements contained in the input information; and performing a registration side replacement process that replaces the order of the elements contained in the input information using a replacement key, and generates the registration information by replacing the order of the elements contained in the secret sharing information using the replacement key.

[0023] A matching information generation program of the present invention is a matching information generation method for generating matching information based on matching side input information input for matching, the matching information including at least one of elements corresponding to a first set and elements corresponding to a second set different from the first set, and the program causes a computer to execute the following steps: generating matching side intermediate information from the matching side input information by performing a matching side duplication process of duplication of elements corresponding to the first set of the input information and a matching side distribution process of secretly sharing elements corresponding to the second set of the input information; and performing a matching side substitution process of replacing the order of elements included in the input information using a substitution key, and generating the matching information by replacing the order of elements included in the matching side intermediate information using the substitution key.

[0024] The matching information generation program of the present invention is a matching information generation program that generates matching information based on matching side input information that is input for matching, and causes a computer to perform a duplication process that duplicates elements contained in the input information to generate duplicate information from the matching side input information, and a matching side replacement process that replaces the order of the elements contained in the input information using a replacement key, thereby generating the matching information by replacing the order of the elements contained in the duplicate information using the replacement key. Effect of the Invention

[0025] According to the present invention, it is possible to provide a registration information generating device, a matching information generating device, a matching system, a registration information generating method, a matching information generating method, a similarity calculation method, a registration information generating program, and a matching information generating program that realize a cancelable conversion that can suppress deterioration of matching accuracy and reduce the risk of information leakage at the same time. Note that the present invention may achieve other effects instead of or in addition to the effect. [Brief description of the drawings]

[0026] [Figure 1] FIG. 1 is a diagram illustrating an example of an operation form of a collation system according to the first embodiment. [Diagram 2] FIG. 2 is a block diagram illustrating a hardware configuration of an information processing device included in the verification system according to the first embodiment. [Diagram 3] FIG. 3 is a diagram showing an overview of cancelable biometrics as a reference example. [Figure 4] FIG. 4 is an explanatory diagram of a cancelable transformation as a reference example. [Diagram 5] FIG. 5 is a functional block diagram illustrating a functional configuration of the registration information generating device according to the first embodiment. [Figure 6] FIG. 6 is a functional block diagram illustrating a functional configuration of the matching information generating device according to the first embodiment. [Figure 7] FIG. 7 is a functional block diagram illustrating the functional configuration of the key generating device according to the first embodiment. [Figure 8] FIG. 8 is a functional block diagram illustrating the functional configuration of the verification device according to the first embodiment. [Figure 9] FIG. 9 is a flowchart showing a flow of a process for generating registration information according to the first embodiment. [Figure 10] FIG. 10 is an explanatory diagram of a process for generating registration information according to the first embodiment. [Figure 11] FIG. 11 is a flowchart showing a process for generating matching information according to the first embodiment. [Figure 12]FIG. 12 is an explanatory diagram of a process for generating matching information according to the first embodiment. [Figure 13] FIG. 13 is a flowchart showing a flow of processing performed by the collation device according to the first embodiment. [Figure 14] FIG. 14 is a functional block diagram illustrating a functional configuration of a registration information generating device according to a modification of the first embodiment. [Figure 15] FIG. 15 is a functional block diagram illustrating a functional configuration of a matching information generating device according to a modification of the first embodiment. [Figure 16] FIG. 16 is a flowchart showing a flow of a process for generating registration information according to the second embodiment. [Figure 17] FIG. 17 is an explanatory diagram of a process for generating registration information according to the second embodiment. [Figure 18] FIG. 18 is a flowchart showing a process for generating matching information according to the second embodiment. [Figure 19] FIG. 19 is an explanatory diagram of a process for generating matching information according to the second embodiment. [Figure 20] FIG. 20 is a diagram showing an operation form of the collation system according to the third embodiment. [Figure 21] FIG. 21 is a functional block diagram illustrating a functional configuration of a registration information generating device according to the third embodiment. [Figure 22] FIG. 22 is a functional block diagram illustrating a functional configuration of a matching information generating device according to the third embodiment. [Figure 23] FIG. 23 is a functional block diagram illustrating a functional configuration of a key generating device according to the third embodiment. [Figure 24] FIG. 24 is a functional block diagram illustrating a functional configuration of a matching device according to the third embodiment. [Diagram 25] FIG. 25 is a flowchart showing a flow of a process for generating registration information according to the third embodiment. [Figure 26] FIG. 26 is an explanatory diagram of a process for generating registration information according to the third embodiment. [Figure 27]FIG. 27 is a flowchart showing a process for generating matching information according to the third embodiment. [Figure 28] FIG. 28 is an explanatory diagram of a process for generating matching information according to the third embodiment. [Figure 29] FIG. 29 is a flowchart showing a flow of processing performed by the collation device according to the third embodiment. [Diagram 30] FIG. 30 is a functional block diagram illustrating a functional configuration of a registration information generating device according to a modification of the third embodiment. [Diagram 31] FIG. 31 is a functional block diagram illustrating a functional configuration of a matching information generating device according to a modification of the third embodiment. [Diagram 32] FIG. 32 is a flowchart showing a flow of a process for generating registration information according to the fourth embodiment. [Diagram 33] FIG. 33 is an explanatory diagram of a process for generating registration information according to the fourth embodiment. [Diagram 34] FIG. 34 is a flowchart showing a process for generating matching information according to the fourth embodiment. [Diagram 35] FIG. 35 is an explanatory diagram of a process for generating matching information according to the fourth embodiment. [Diagram 36] FIG. 36 is a model diagram of a cancelable transformation for Euclidean distance according to the second reference example. [Figure 37] FIG. 37 is a diagram illustrating an example of a modified form of information according to the present invention. [Figure 38] FIG. 38 is a block diagram illustrating an outline of a verification system according to the fifth embodiment. [Figure 39] FIG. 39 is a block diagram illustrating an outline of a registration information generating device according to the fifth embodiment. [Diagram 40] FIG. 40 is a block diagram illustrating an outline of a matching information generating device according to the fifth embodiment. [Diagram 41] FIG. 41 is a block diagram illustrating an outline of a verification system according to the sixth embodiment. [Diagram 42]FIG. 42 is a block diagram illustrating an outline of a registration information generating device according to the sixth embodiment. [Diagram 43] FIG. 43 is a block diagram illustrating an outline of a matching information generating device according to the sixth embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0027] Hereinafter, an embodiment of the present invention will be described in detail with reference to the accompanying drawings. In this specification and the drawings, elements that can be similarly described may be designated by the same reference numerals to avoid repetitive description.

[0028] Each embodiment described below is merely one example of a configuration that can realize the present invention. Each of the following embodiments can be modified or changed as appropriate depending on the configuration of the device to which the present invention is applied and various conditions. Not all of the combinations of elements included in each of the following embodiments are necessarily essential to realize the present invention, and some of the elements can be omitted as appropriate. Therefore, the scope of the present invention is not limited by the configurations described in each of the following embodiments. As long as there is no mutual contradiction, a configuration that combines multiple configurations described in the embodiments can also be adopted.

[0029] The explanation will be given in the following order: 1. Overview of the embodiments of the present invention 2. First embodiment 2.1. Overview of the operation of the matching system 2.2.Hardware configuration of information processing device 2.3. Overview of Cancellable Biometrics 2.4. Functional configuration of the registration information generating device 2.5. Functional configuration of the matching information generation device 2.6. Functional configuration of the key generation device 2.7. Functional configuration of the verification device 2.8. Process flow for generating registration information 2.9. Process flow for generating matching information 2.10.Process flow for calculating similarity 2.11. Modifications of the Registration Information Generation Device and the Matching Information Generation Device 3. Second embodiment 3.1. Process flow for generating registration information 3.2. Process flow for generating matching information 3.3. Calculation of Similarity in the Second Embodiment 4. Third embodiment 4.1. Overview of the operation of the matching system 4.2. Functional configuration of the registration information generating device 4.3. Functional configuration of the matching information generation device 4.4. Functional configuration of the key generation device 4.5. Functional configuration of the verification device 4.6. Process flow for generating registration information 4.7. Process flow for generating matching information 4.8.Process flow for calculating similarity 4.9. Modifications of the Registration Information Generation Device and the Matching Information Generation Device 5. Fourth embodiment 5.1. Process flow for generating registration information 5.2. Process flow for generating matching information 5.3.Process flow for calculating similarity 6. Application to Euclidean distance 7. Fifth embodiment 8. Sixth embodiment 9. Other embodiments

[0030] <1. Overview of the embodiment of the present invention> First, an outline of the embodiment of the present invention will be described.

[0031] (1)Technical issues In recent years, biometric authentication technology has been widely used as one of user authentication technologies, which performs authentication by matching biometric information such as fingerprints, veins, irises, and facial images. In biometric authentication technology, features are extracted from the biometric information of a user and stored as registered information, and the features of the biometric information of the user input during authentication are matched against the registered information to determine whether or not the user can be authenticated.

[0032] Compared with authentication technologies using IC cards, etc., biometric authentication technology has the advantage of being less prone to loss or theft, but even if biometric information is leaked, it cannot be discarded or updated. In other words, the leakage of biometric information not only poses a problem regarding personal information of the living body related to the leaked biometric information, but also poses a problem of compromising the security of authentication systems that use the biometric information of the living body related to the leaked biometric information.

[0033] To address these issues, a technology called "cancellable biometrics" has been proposed that makes it possible to invalidate registered information in order to protect the biometric information of users.

[0034] In cancelable biometrics, the elements of the feature extracted from the biometric information are encrypted by rearranging them using a secret key to generate registration information. In addition, the elements of the feature extracted from the biometric information input during authentication are also encrypted using the same secret key as used during registration to generate matching information. In other words, when recreating the registration information, it is possible to invalidate the already generated registration information by using a different secret key. The encryption in cancelable biometrics is called "cancellable transformation" due to the nature of such registration information. As a cancelable transformation, for example, a method has been proposed in which an image of feature points obtained from a fingerprint or the like is divided into blocks, and the order of the divided images is rearranged using a secret key to generate registration information and matching information.

[0035] As described above, in the cancelable transformation, the elements of the feature extracted from the biometric information are transformed using the same secret key at the time of enrollment and at the time of matching. Therefore, in order to reduce the risk of leakage of the biometric information, it is desirable to match the enrollment information with the matching information while keeping the information in the cancelable transformed state. In addition, when matching the enrollment information with the matching information, it is desirable to perform the cancelable transformation so that matching accuracy equivalent to that of conventional biometric authentication technology can be obtained.

[0036] In consideration of the above circumstances, this embodiment provides a registration information generating device, a matching information generating device, a matching system, a registration information generating method, a matching information generating method, a similarity calculation method, a registration information generating program, and a matching information generating program that realizes a cancellable conversion that can simultaneously suppress deterioration of matching accuracy and reduce the risk of information leakage.

[0037] (2) Technical features In an embodiment of the present invention, a registration information generation device generates registration information based on registration side input information input for registration, wherein the registration side input information includes at least one of an element corresponding to a first set and an element corresponding to a second set different from the first set, and the registration information generation device is equipped with a registration side intermediate information generation unit that generates registration side intermediate information from the registration side input information by performing a registration side distribution process that secretly distributes elements corresponding to the first set of the input information and a registration side duplication process that duplicates elements corresponding to the second set of the input information, and a registration side replacement processing unit that performs a registration side replacement process that replaces the order of elements included in the input information using a replacement key, and the registration information is generated by replacing the order of elements included in the registration side intermediate information using the replacement key.

[0038] In another embodiment of the present invention, a registration information generation device generates registration information based on registration side input information input for registration, and includes a distribution processing unit that generates secret sharing information from the registration side input information by performing a distribution process that secretly distributes elements included in the input information, and a registration side replacement processing unit that performs a registration side replacement process that replaces the order of the elements included in the input information using a replacement key, and the registration information is generated by replacing the order of the elements included in the secret sharing information using the replacement key.

[0039] This makes it possible to provide a registration information generating device, a matching information generating device, a matching system, a registration information generating method, a matching information generating method, a similarity calculation method, a registration information generating program, and a matching information generating program that realize a cancelable conversion that can simultaneously suppress deterioration of matching accuracy and reduce the risk of information leakage. Note that the above-mentioned technical features are specific examples of embodiments of the present invention, and naturally, embodiments of the present invention are not limited to the above-mentioned technical features.

[0040] 2. First embodiment A first embodiment of the present invention will be described below with reference to Fig. 1 to Fig. 15. In this embodiment, a matching system using biometric information of a user such as a fingerprint, vein, iris, and face image will be described.

[0041] <2.1. Overview of the operation of the matching system> Fig. 1 is a diagram showing an example of an operation form of a matching system 1000A according to a first embodiment of the present invention. As shown in Fig. 1, the matching system 1000A is configured by connecting a registration information generating device 1A, a matching information generating device 2A, a key generating device 3A, a matching device 4A, and a database (hereinafter sometimes referred to as "DB") 5A via a network 6.

[0042] The registration information generating device 1A is an information processing device such as a computer or a server in which a program is installed that generates registration information to be registered in the verification system 1000A from a user's biometric information. The registration information generating device 1A is realized by, for example, a portable information processing terminal such as a smartphone, an ATM (Automatic Teller Machine), a PC (Personal Computer) connected to a sensor that detects biometric information, or the like. The process of generating registration information in the registration information generating device 1A will be described in detail later.

[0043] The matching information generating device 2A is an information processing device such as a computer or a server in which a program is installed that generates matching information from a user's biometric information for matching with registered information registered in the matching system 1000A. The matching information generating device 2A is realized by, for example, a portable information processing terminal such as a smartphone, an ATM, a PC connected to a sensor that detects biometric information, or the like. The process of generating matching information in the matching information generating device 2A will be described in detail later.

[0044] The key generating device 3A is an information processing device that generates a private key used when converting a user's biometric information into registration information or matching information. The key generating device 3A generates, for example, a random number, a mask array, and a substitution key as the private key. The key generating device 3A transmits the generated private key to the registration information generating device 1A and the matching information generating device 2A via the network 6.

[0045] The matching device 4A is an information processing device that matches matching information with registered information. For example, in biometric authentication technology, the matching device 4A matches the feature amount of the biometric information of the user stored as registered information with the feature amount of the biometric information of the user input for authentication. The matching system 1000A may authenticate the user identified by the matching information based on the result of matching the matching information with the registered information.

[0046] The DB5A is a storage medium that stores the registration information generated by the registration information generating device 1A, and the verification device 4A collates the registration information stored in the DB5A with the verification information. Although the verification device 4A and the DB5A are shown as independent elements in Fig. 1, the DB5A may be implemented in a storage medium such as a hard disk drive (HDD) included in the verification device 4A.

[0047] In addition, while FIG. 1 shows an example in which the registration information generating device 1A and the matching information generating device 2A are connected to the network 6, they do not necessarily need to be connected to the network 6. For example, the registration information generating device 1A or the matching information generating device 2A may be connected to the key generating device 3A, the matching device 4A, and the DB 5A by a USB (Universal Serial Bus) or the like. Furthermore, the registration information generating device 1A may be connected to the DB 5A when transmitting the registration information to the DB 5A. Furthermore, the matching information generating device 2A may be connected to the matching device 4A when transmitting the matching information. Furthermore, the key generating device 3A may be connected to the registration information generating device 1A or the matching information generating device 2A when transmitting the private key.

[0048] <2.2. Hardware configuration of information processing device> Next, the hardware configurations of information processing devices, such as a registration information generating device 1A, a matching information generating device 2A, a key generating device 3A, and a matching device 4A according to this embodiment, will be described with reference to Fig. 2. Fig. 2 is a block diagram showing the hardware configuration of the information processing device.

[0049] The information processing device includes a central processing unit (CPU) 11, a read only memory (ROM) 12, a random access memory (RAM) 13, a storage medium 14, and an interface (I / F) 15, which are connected via a bus. 21 The I / F 15 is connected to an input unit 17, a display unit 18, a sensor 19, and the network 6. The sensor 19 may be omitted.

[0050] The CPU 11 is a calculation means and controls the operation of the entire information processing device. The RAM 13 is a volatile storage medium capable of reading and writing information at high speed, and is used as a working area when the CPU 11 processes information. The ROM 12 is a read-only non-volatile storage medium in which programs such as firmware are stored. The storage medium 14 is a non-volatile storage medium such as a HDD in which information can be read and written, and in which information such as the OS (Operating System), various control programs, and application programs is stored.

[0051] I / F15 is a bus 21 and connects and controls various hardware and the network 6. The input unit 17 is an input device such as a keyboard or a mouse that allows the user to input information to the information processing device. The display unit 18 is a display device such as an LCD (Liquid Crystal Display) that allows the user to check the status of the information processing device. Note that the input unit 17 and the display unit 18 are optional.

[0052] The sensor 19 is a module that acquires biometric information of a user, such as a fingerprint, vein, face image, etc. The sensor 19 may be omitted. In this case, the registration information generating device 1A and the matching information generating device 2A may acquire the biometric information of the user via the network 6.

[0053] In such a hardware configuration, the CPU 11 of the registration information generating device 1A performs calculations according to a program stored in the ROM 12 of the registration information generating device 1A or a program loaded from the storage medium 14 of the registration information generating device 1A into the RAM 13 of the registration information generating device 1A, thereby forming a software control unit of the registration information generating device 1A.

[0054] A functional block that realizes the functions of the controller 100A (see FIG. 5) of the registration information generating device 1A according to this embodiment is configured by combining the software control unit configured as described above with hardware.

[0055] Furthermore, in the above-described hardware configuration, a software control unit of the matching information generating device 2A is configured by the CPU 11 of the matching information generating device 2A performing calculations in accordance with a program stored in the ROM 12 of the matching information generating device 2A or a program loaded from the storage medium 14 of the matching information generating device 2A to the RAM 13 of the matching information generating device 2A.

[0056] A functional block that realizes the functions of the controller 200A (see FIG. 6) of the matching information generation device 2A according to this embodiment is configured by combining the software control unit configured as described above with hardware.

[0057] Furthermore, in the above-described hardware configuration, a software control unit of the key generation device 3A is configured by the CPU 11 of the key generation device 3A performing calculations in accordance with a program stored in the ROM 12 of the key generation device 3A or a program loaded from the storage medium 14 of the key generation device 3A to the RAM 13 of the key generation device 3A.

[0058] A functional block that realizes the functions of the controller 300A (see FIG. 7) of the key generation device 3A according to this embodiment is configured by combining the software control unit configured as described above with hardware.

[0059] In addition, in the above-mentioned hardware configuration, the CPU 11 of the matching device 4A performs calculations according to a program stored in the ROM 12 of the matching device 4A or a program loaded from the storage medium 14 of the matching device 4A to the RAM 13 of the matching device 4A, thereby forming a software control unit of the matching device 4A.

[0060] A functional block that realizes the functions of the controller 400A (see FIG. 8) of the verification device 4A according to this embodiment is configured by combining the software control unit configured as described above with hardware.

[0061] <2.3. Overview of Cancellable Biometrics> Next, an overview of cancelable biometrics that can invalidate registered information will be described with reference to Fig. 3 and Fig. 4. Fig. 3 is a diagram showing an overview of cancelable biometrics. Fig. 4 is an explanatory diagram of an information conversion algorithm in cancelable biometrics.

[0062] In biometric authentication technology, a user's biometric information, such as a fingerprint, vein, iris, and face image, is registered in advance, and authentication is determined based on the result of comparing the biometric information of the user input during personal authentication. However, if biometric information is leaked, not only is there a problem regarding personal information of the living body related to the leaked biometric information, but there is also a problem that the security of the authentication system using the biometric information of the living body related to the leaked biometric information is impaired.

[0063] To address these issues, a method known as "cancellable biometrics" is being used to protect users' biometric information. This method uses registered information that keeps the biometric information secret, and makes it possible to invalidate the leaked registered information if it is leaked.

[0064] Here, an overview of cancelable biometrics will be described with reference to FIG. 3. First, when biometric information is registered, features are extracted from the biometric information. For the sake of explanation, feature data, which is a vector, is extracted from the biometric information.

number

number

number

[0065] Next, the process during authentication will be described. During authentication, features are extracted from biometric information. For the sake of explanation, it is assumed here that feature data, which is a vector, is extracted from the biometric information input for authentication.

number

number

[0066] As shown in (Equation 1-1) and (Equation 1-2), in cancelable biometrics, features are converted by a conversion function F using the same conversion key K at the time of enrollment and authentication. In other words, when feature data x and feature data y close to feature data x are extracted from the fingerprint of the middle finger of the right hand of the same person as biometric information, enrollment data T and matching data T' are generated by converting feature data x and feature data y that are close to each other, respectively. As a result, it is determined that matching data T' matches enrollment data T.

[0067] Here, for the same feature data x, the conversion key

number

number

number

number

[0068] As described above, the conversion function F is a function that performs one-way conversion on input data. In other words, when the conversion key K1 is not equal to the conversion key K2, the order of elements of the data T1 output by the conversion function F when the conversion key K1 is used for the feature data x is different from the order of elements of the data T2 output by the conversion function F when the conversion key K2 is used. As described above, in the cancelable biometrics, information on the conversion key K1 and the conversion key K2 is not required for matching.

[0069] When data T2 is compared against data T1, it is determined that data T1 and data T2 do not match, because both data cannot be identified as having been generated from feature data x, even though they were generated by converting feature data x.

[0070] In this way, in the cancelable biometrics, different data T1 and data T2 are generated by using different conversion keys K1 and K2 for the same data. By using such a characteristic, in the cancelable biometrics, the registration information stored in the database or the like can be invalidated by changing the conversion key K. In the following description, the data conversion method in the cancelable biometrics may be referred to as "cancelable conversion".

[0071] Next, application of cancelable transformation to normalized correlation will be described with reference to Fig. 4. Fig. 4 is an explanatory diagram of cancelable transformation to normalized correlation as a reference example.

[0072] In face recognition technology that uses a user's facial image as biometric information, the features extracted from the biometric information are converted into feature data, which is an n-dimensional vector.

number

[0073] Feature data X ref When generating registration information for face recognition from ref For each dimension of the elements, the mask array

number

number

[0074] Next, the feature data X after positive / negative conversion ref The order of elements in each dimension of is the permutation key

number

number

[0075] Next, in face recognition, when generating matching information for matching, the features extracted from the biometric information used for matching are converted into feature data, which is an n-dimensional vector.

number

number

[0076] And the feature data X ref and feature data Y ref The similarity between feature data X and ref and feature data Y ref Normalized correlation with

number

number

[0077] Then, for any positive integers i, j,

number

number

[0078] By the way, as a method of performing secret calculations while encrypting information to be concealed, such as biometric information, homomorphic encryption, which performs calculations while both the program and data are encrypted, is known. However, homomorphic encryption, which performs calculations while both the program and data are encrypted, requires more calculation time than calculations using plain text, and is therefore not suitable for systems that require high throughput. Compared to homomorphic encryption, cancelable biometrics has a lighter processing load, and is therefore suitable for systems that require high throughput.

[0079] However, in an authentication system using cancelable biometrics, the mask array R ref Masking and substitution key K ref Therefore, if an attacker finds data that is close to feature data X,

number

[0080] In addition, the registration data T ref If the registration data T refThe same mask array R used to generate ref and replacement key K ref Using the above, the registration data T ref New biometric feature data X of the same person related to ref 2 to registration data T ref Let's assume that you generate 2 and register it in the authentication system.

[0081] In this case, the registration data T ref Feature data X relating to ref And feature data X ref Since the data is close to 2, the registered data T ref 2, the leaked registration data T ref However, there is a problem in that it is easy to carry out so-called spoofing, which involves performing matching using a unique identifier.

[0082] Furthermore, registration information T ref In case that is leaked, in order to ensure the security of the entire authentication system, the mask array R ref and replacement key K ref It is necessary to update the registration data T ref into a new mask array R ref and replacement key K ref There is a problem in that the registration information must be updated to the one generated using

[0083] In response to such issues, this embodiment describes a registration information generating device, a matching information generating device, a matching system, a registration information generating method, a matching information generating method, a similarity calculation method, a registration information generating program, and a matching information generating program that realizes a cancellable conversion that can simultaneously suppress deterioration of matching accuracy and reduce the risk of information leakage.

[0084] <2.4. Functional configuration of the registration information generating device> Next, the functional configuration of the registration information generating device 1A will be described with reference to Fig. 5. Fig. 5 is a functional block diagram illustrating the functional configuration of the registration information generating device 1A. As shown in Fig. 5, the registration information generating device 1A includes a controller 100A and a network I / F 101A which is an interface for the controller 100A to exchange information with other devices via a network 6.

[0085] The controller 100A generates registration information used in the matching system 1000A from information acquired via the network I / F 101A. The registration information is one of the pieces of information used for matching in the matching system 1000A. The registration information generated by the registration information generating device 1A is stored in the DB5A. The controller 100A is realized by installing a dedicated software program in an information processing device such as the registration information generating device 1A. The controller 100A includes a feature extraction unit 110A, a distribution processing unit 120A, a replacement processing unit 130A, and a mask processing unit 140A.

[0086] During authentication, the feature extraction unit 110A extracts features from the biometric information input to the controller 100A via the sensor 19 or I / F 15 of the registration information generating device 1A to generate feature data. In this embodiment, the feature data X, which is an n-dimensional vector, is generated from biometric information such as a fingerprint, vein, iris, and face image.

[0087] The distribution processing unit 120A performs a distribution process for secretly sharing the input information elements. The details of the distribution process in this embodiment will be described later.

[0088] The substitution processing unit 130A performs substitution processing for substituting the arrangement of elements of input information based on a private key such as a substitution key K. Details of the substitution processing in this embodiment will be described later.

[0089] The mask processing unit 140A performs mask processing to convert the elements of the input information using a mask array. The details of the mask processing in this embodiment will be described later.

[0090] <2.5. Functional configuration of the matching information generating device> Next, the functional configuration of the matching information generating device 2A will be described with reference to Fig. 6. Fig. 6 is a functional block diagram illustrating the functional configuration of the matching information generating device 2A. As shown in Fig. 6, the matching information generating device 2A includes a controller 200A and a network I / F 201A that is an interface for the controller 200A to exchange information with other devices via a network 6.

[0091] The controller 200A generates matching information used in the matching system 1000A from information acquired via the network I / F 201A. The matching information is one of the pieces of information used for matching in the matching system 1000A. The matching information generated by the matching information generating device 2A is transmitted to the matching device 4A. The controller 200A is realized by installing a dedicated software program in an information processing device such as the matching information generating device 2A. The controller 200A includes a feature extraction unit 210A, a duplication processing unit 220A, a replacement processing unit 230A, and a mask processing unit 240A.

[0092] During authentication, the feature extraction unit 210A extracts features from biometric information input to the controller 200A via the sensor 19 or I / F 15 of the matching information generation device 2A. In this embodiment, feature data Y, which is an n-dimensional vector, is extracted from biometric information such as a fingerprint, vein, iris, and face image.

[0093] The duplication processing unit 220A performs duplication processing for duplication of elements of input information. Details of the duplication processing in this embodiment will be described later.

[0094] The substitution processing unit 230A performs substitution processing to substitute the arrangement of elements of input information based on a private key such as a substitution key K. Details of the substitution processing in this embodiment will be described later.

[0095] The mask processing unit 240A performs mask processing to convert the elements of the input information using a mask array. The details of the mask processing in this embodiment will be described later.

[0096] <2.6. Functional configuration of the key generation device> Next, the functional configuration of the key generation device 3A will be described with reference to Fig. 7. Fig. 7 is a functional block diagram illustrating the functional configuration of the key generation device 3A. As shown in Fig. 7, the key generation device 3A includes a controller 300A and a network I / F 301A that is an interface for the controller 300A to exchange information with other devices via a network 6.

[0097] The controller 300A generates a private key used in the verification system 1000A from information acquired via the network I / F 301A. The controller 300A is realized by installing a dedicated software program in an information processing device such as the key generating device 3A. The controller 300A includes an index information generating unit 310A, a substitution key generating unit 320A, and a mask array generating unit 330A.

[0098] The index information generating unit 310A generates index information, which is information for identifying what processing is to be performed on each element included in the information. As described above, the registration information generating device 1A performs distributed processing. The index information generating unit 310A generates, as the index information, information indicating the elements to be subjected to distributed processing, for example.

[0099] Furthermore, as described above, the matching information generating device 2A performs the duplication process. The index information generating unit 310A generates, as the index information, for example, information indicating the element to be subjected to the duplication process. The network I / F 301A transmits the index information generated by the index information generating unit 310A to the registration information generating device 1A and the matching information generating device 2A.

[0100] The replacement key generating unit 320A generates private keys such as the replacement key K used by the registration information generating device 1A and the matching information generating device 2A in the replacement process. The network I / F 301A transmits the private keys such as the replacement key K generated by the replacement key generating unit 320A to the registration information generating device 1A and the matching information generating device 2A.

[0101] The mask sequence generation unit 330A generates a mask sequence used in the mask process by the registration information generation device 1A and the matching information generation device 2A. The network I / F 301A transmits the mask sequence generated by the mask sequence generation unit 330A to the registration information generation device 1A and the matching information generation device 2A.

[0102] <2.7. Functional configuration of the verification device> Next, the functional configuration of the collation device 4A will be described with reference to Fig. 8. Fig. 8 is a functional block diagram illustrating the functional configuration of the collation device 4A. As shown in Fig. 8, the collation device 4A includes a controller 400A and a network I / F 401A that is an interface for the controller 400A to exchange information with other devices via a network 6.

[0103] The controller 400A acquires information input to the matching system 1000A via the network I / F 401A and performs matching based on the acquired information. The controller 400A is realized by installing a dedicated software program in an information processing device such as the matching device 4A. The controller 400A includes a similarity calculation unit 410A and a matching processing unit 420A.

[0104] The similarity calculation unit 410A calculates a similarity indicating how similar each piece of information is for a plurality of pieces of information. In this embodiment, the similarity calculation unit 410A calculates the similarity based on a normalized correlation between the registered information and the matching information. Details of the process performed by the similarity calculation unit 410A will be described later.

[0105] The matching processing unit 420A performs a matching process to determine whether or not a plurality of pieces of information match each other based on the similarity calculated by the similarity calculation unit 410A. For example, when the similarity calculation unit 410A calculates the similarity between the registered information and the matching information, the matching processing unit 420A performs the matching process to determine whether or not the matching information matches the registered information based on the similarity calculated by the similarity calculation unit 410A.

[0106] With the configuration described above, the verification system 1000A calculates the similarity between the registration information generated by the registration information generating device 1A and the verification information generated by the verification information generating device 2A. Then, based on the calculated similarity, it is determined whether the registration information and the verification information match each other. Note that information on the determination result indicating whether the verification information in the verification process matches the registration information may be used as information for determining whether authentication is possible in an authentication system having the verification system 1000A.

[0107] <2.8. Process flow for generating registration information> Next, a process flow when the registration information generating device 1A performs cancelable conversion of input information in this embodiment will be described. In the verification system 1000A, when performing verification to determine whether or not biometric authentication is possible, first, registration information, which is one of the pieces of information used for verification, is generated from the biometric information of the user input to the verification system 1000A. The registration information generating device 1A acquires the biometric information of the user input to the verification system 1000A via the sensor 19 or I / F 15, and generates registration information based on the acquired biometric information.

[0108] Fig. 9 is a flowchart showing the flow of a process for generating registration information in the registration information generating device 1A. Fig. 10 is an explanatory diagram modeling the process for generating registration information in the registration information generating device 1A. The flow of cancelable conversion in the registration information generating device 1A will be described with reference to Figs. 9 and 10.

[0109] The biometric information of the user acquired by the registration information generating device 1A is input to the controller 100A via the network I / F 101A. In the explanation of Fig. 9 and Fig. 10, it is assumed that the facial image information of the user A is input to the registration information generating device 1A.

[0110] In step S11, the feature extraction unit 110A extracts feature data from the face image information of the user A.

number

[0111] Next, in step S12, the distribution processing unit 120A performs a distribution process to distribute the first-dimension element to the n-dimension element of the n-dimensional feature amount data X. For example, the distribution processing unit 120A distributes the first-dimension element x1 of the n-dimensional feature amount data X as follows:

number

number

number

[0112] The distributed processing of this embodiment corresponds to a process in which the distributed processing unit 120A distributes each element of each dimension of the feature amount data X to a plurality of random numbers. i is distributed into multiple random numbers (in equation 2-2, random numbers x1[1] and x1[2]).

number

number

[0113] As shown in (Equation 2-2), the random number x i [1] and a random number x i [2] is a random number x i [1] and a random number x i [2] is added to the original element x i Therefore, the random number x i [1] and a random number x i [2] is the random number x i [1] and a random number x i[2] By itself, the i-th element x of the original n-dimensional feature data X is i In this embodiment, the i-th element x of the feature amount data X is i The i-th element x of the feature data X alone i In some cases, dividing a secret into multiple elements (for example, random numbers x1[1] and x1[2] in (Equation 2-2)) that cannot be restored to their original state is referred to as “secret sharing.”

[0114] In (Equation 2-2), the distributed processing unit 120A calculates the i-th element x i The secret is shared to generate two random numbers (random number x i [1] and a random number x i [2]), but the i-th element x of the n-dimensional feature data x i The number of divisions is not limited to 2. For example, the distributed processing unit 120A divides the i-th element x of the n-dimensional feature amount data x into i Let the number of divisions be j (i, j are any positive integers).

number

number

[0115] Next, in step S13, the mask processing unit 140A applies a mask array

number

number

[0116] Next, in step S14, the substitution processing unit 130A substitutes the elements of each dimension of the positive / negative converted shared data with a substitution key

number

number

[0117] For example, the elements of each dimension included in the distribution data after positive / negative conversion of the registered data T (i, j are any positive integers) are

number

[0118] The registration data T may be generated by omitting the masking process using the mask array R. In this case, the data T, which is a vector generated as a result of performing the replacement process by the replacement processor 130A on the distributed data, is an example of the registration information. Furthermore, the registration data T may be generated by switching the order of the masking process and the replacement process.

[0119] The registration information generating device 1A transmits the thus generated registration data T to the DB 5A in step S15. The DB 5A stores the received registration data T. Note that in step S15, the registration information generating device 1A may transmit the registration data T related to the biometric information of the user A to the DB 5A together with an identifier (user identifier) ​​that can identify the registration data T as information related to the user A.

[0120] In this way, the registration information generating device 1A performs masking and replacement processing after secretly sharing the elements included in the feature data X to generate the registration data T. Therefore, if the replacement key K is not leaked, it is difficult to determine which element of the feature data X corresponds to the element included in the registration data T. In addition, since the registration data T is generated after secretly sharing the feature data X, even if both the mask array R and the replacement key K are leaked, it is difficult to distinguish that the registration data newly generated using the leaked mask array R and replacement key K and the leaked registration data T are derived from the same biometric feature data. As described above, the registration information generating device 1A of this embodiment can further reduce the risk of information leakage.

[0121] <2.9. Process flow for generating matching information> Next, a process flow when the matching information generating device 2A performs cancelable conversion of input information in this embodiment will be described. After storing the registration information in DB5A, the matching system 1000A needs to use biometric information of the same living body as the registration information when performing matching to determine whether or not the biometric authentication can be performed. The matching information generating device 2A acquires the biometric information of the user via the sensor 19 or I / F 15, and performs cancelable conversion based on the acquired biometric information. Information generated by cancelable conversion of the biometric information by the matching information generating device 2A is an example of matching information to be matched against the registration information stored in DB5A.

[0122] Fig. 11 is a flowchart showing the flow of processing for generating matching information in the matching information generation device 2A. Fig. 12 is an explanatory diagram modeling the processing for generating matching information in the matching information generation device 2A. The flow of cancelable conversion in the matching information generation device 2A will be described with reference to Figs. 11 and 12.

[0123] The biometric information of the user acquired by the matching information generation device 2A is input to the controller 200A via the network I / F 201A. In the explanation of Fig. 11 and Fig. 12, it is assumed that facial image information of the user A is input to the matching information generation device 2A.

[0124] In step S21, the feature extraction unit 210A extracts feature data from the face image information of the user A.

number

[0125] Next, in step S22, the duplication processing unit 220A performs duplication processing to duplicate elements from the first dimension to the nth dimension of the n-dimensional feature amount data Y. As a result, the elements of each dimension of the n-dimensional feature amount data Y are doubled, and data of a 2n-dimensional vector is obtained as shown in (Equation 3-2).

number

[0126] The duplication process of this embodiment corresponds to a process in which the duplication processor 220A duplicates elements of each dimension of the feature amount data Y. In addition, in the registration information generating device 1A, when elements of each dimension of the feature amount data X are divided into a division number n1, the duplication processor 220A performs duplication process to multiply the elements of each dimension of the feature amount data Y by n2. Here, n1=n2, and both n1 and n2 are integers equal to or greater than 2. In the following description, 2n-dimensional vector data Y generated by duplication process of the n-dimensional feature amount data Y is referred to ascp This may be referred to as “post-duplication data.” The post-duplication data is an example of duplication information in this embodiment.

[0127] In this embodiment, distribution processing and replication processing are performed so that the number of divisions of the feature amount data X in the registration information generating device 1A is the same as the number of copies of the feature amount data Y in the matching information generating device 2A.

[0128] Next, in step S23, the mask processing unit 240A performs mask processing on the elements of each dimension of the copied data using the mask array R (see equation 2-5) (see the third row from the top of FIG. 12). As described above, the mask array R is information generated by the key generation device 3A. The key generation device 3A transmits the mask array R, which is a 2n-dimensional vector, to the matching information generation device 2A. In this embodiment, the key generation device 3A transmits the same mask array R to the registration information generation device 1A and the matching information generation device 2A.

[0129] The positive and negative signs of the elements of each dimension of the duplicated data can be converted by a mask process of multiplying the elements of each dimension of the duplicated data by the mask array R. In the following description, the duplicated data after the mask process of multiplying the elements of each dimension by the mask array R may be referred to as "positive / negative converted duplicated data."

[0130] Next, in step S24, the substitution processing unit 230A performs a substitution process to rearrange elements of each dimension of the positive / negative converted duplicated data by a substitution function F using a substitution key K. The substitution key K is an example of information that defines how to rearrange elements included in the target information, and corresponds to an encryption key randomly generated by the key generating device 3A. As described above, the substitution key K is information generated by the key generating device 3A, and is information transmitted from the key generating device 3A to the matching information generating device 2A in order to generate matching information. In this embodiment, the key generating device 3A transmits the same substitution key K to the registration information generating device 1A and the matching information generating device 2A.

[0131] The substitution processing unit 230A performs substitution processing to convert the positive / negative converted duplicated data into

number

[0132] The same substitution key K is used in the registration information generating device 1A and the matching information generating device 2A. That is, in the substitution process performed by the matching information generating device 2A, the elements of each dimension included in the positive / negative converted duplicated data are rearranged in the same order as in the substitution process in the registration information generating device 1A.

[0133] Therefore, in the collation data T' (i, j are any positive integers) generated in step S24, the elements of each dimension included in the positive / negative converted duplicated data are

number

[0134] The matching data T' may be generated without performing the mask process using the mask array R. In this case, the data T', which is a vector generated as a result of performing the replacement process by the replacement processor 230A on the post-duplicate data, is Matching Further, the order of the mask process and the replacement process may be reversed to generate the matching data T'.

[0135] Next, in step S25, the network I / F 201A transmits the matching data T' to the matching device 4A. At this time, the network I / F 201A may transmit an identifier (user identifier) ​​that can be identified as information related to user A together with the matching data T' to the matching device 4A.

[0136] In this way, the matching information generation device 2A generates matching data T' by duplicating the elements included in the feature data Y and then performing masking and replacement processes. Therefore, if the replacement key K is not leaked, it is difficult to determine which element of the feature data Y corresponds to an element included in the matching data T'. As described above, the matching information generation device 2A of this embodiment can further reduce the risk of information leakage.

[0137] <2.10. Process flow for calculating similarity> Next, a process flow in the matching device 4A in this embodiment will be described with reference to Fig. 13. Fig. 13 is a flowchart showing a process flow in the matching device 4A for matching biometric information input to the registration information generating device 1A with biometric information input to the matching information generating device 2A.

[0138] In step S31, the network I / F 401A acquires the enrollment data T from the DB 5A. The DB 5A stores the enrollment data T related to the biometric information of the user A received from the enrollment information generating device 1A in step S15 together with an identifier that can identify the information as information related to the user A. The matching device 4A acquires, from the DB 5A, the enrollment data T related to the biometric information of the user corresponding to the user identifier received from the matching information generating device 2A in step S25, that is, the user A.

[0139] Next, in step S32, the similarity calculation section 410A calculates the similarity between the feature amount data X and the feature amount data Y.

[0140] In step S32, the similarity calculation unit 410A calculates the inner product of the enrollment data T and the matching data T' as a normalized correlation between the enrollment data T and the matching data T'. Here, the inner product of the enrollment data T and the matching data T', each of which is a 2n-dimensional vector,

number

number

number

[0141] That is, the similarity calculation section 410A calculates the inner product of the registered data T and the matching data T′ as the similarity between the feature amount data X and the feature amount data Y.

[0142] Next, in step S33, the matching processing unit 420A performs a matching process to match the biometric information related to the feature data X with the biometric information related to the feature data Y based on the similarity between the feature data X and the feature data Y calculated in step S32.

[0143] The verification device 4A may authenticate the user identified by the verification information based on the result of the verification process in step S33.

[0144] In addition, for the mask array R used in the registration information generating device 1A, the matching information generating device 2A uses a mask array

number

number

number

[0145] In this way, in the calculation process using the enrollment data T generated by the enrollment information generating device 1A according to this embodiment and the matching data T' generated by the matching information generating device 2A, it is possible to calculate the similarity by calculating the inner product once, so that it is possible to expect a high throughput similar to that of conventional cancelable biometrics. In addition, since the inner product of the enrollment data T and the matching data T' is equal to the inner product of the feature data X and the feature data Y, it is possible to maintain the same matching accuracy as that of conventional cancelable biometrics. Therefore, the enrollment information generating device 1A and the matching information generating device 2A according to this embodiment can realize a cancelable conversion that can simultaneously suppress deterioration of the matching accuracy and reduce the risk of information leakage.

[0146] <2.11. Modifications of the registration information generating device and the matching information generating device> The registration information generating device 1A and the matching information generating device 2A may each include an element that generates a mask array R and a substitution key K. Fig. 14 is a functional block diagram showing a functional configuration of the registration information generating device 1A according to a modification of the first embodiment. Fig. 15 is a functional block diagram showing a functional configuration of the matching information generating device 2A according to a modification of the first embodiment.

[0147] As shown in FIG. 14, the registration information generating device 1A includes a feature extracting unit 110A, a distribution processing unit 120A, a substitution processing unit 130A, and a mask processing unit 140A (see FIG. 5), as well as a substitution key generating unit 150A and a mask sequence generating unit 160A.

[0148] The substitution key generating unit 150A generates a private key such as a substitution key K used by the registration information generating device 1A in the substitution process. The mask sequence generating unit 160A generates a mask sequence used by the registration information generating device 1A in the mask process.

[0149] In addition to what is shown in FIG. 14, the registration information generating device 1A may also include an element equivalent to an index information generating unit 310A that generates index information, which is information for identifying what processing should be performed on each element contained in the information.

[0150] As shown in FIG. 15, the matching information generation device 2A includes a feature extraction unit 210A, a duplication processing unit 220A, a substitution processing unit 230A, and a mask processing unit 240A (see FIG. 6), as well as a substitution key generation unit 250A and a mask sequence generation unit 260A.

[0151] The substitution key generation unit 250A generates a private key such as the substitution key K used by the matching information generation device 2A in the substitution process. The mask sequence generation unit 260A generates a mask sequence used by the matching information generation device 2A in the mask process.

[0152] In addition to what is shown in FIG. 15, the matching information generating device 2A may also include an element equivalent to an index information generating unit 310A that generates index information, which is information for identifying what processing should be performed on each element contained in the information.

[0153] According to this modification, the registration information generating device 1A and the matching information generating device 2A can perform the cancelable conversion without communicating with the key generating device 3A. Therefore, it is possible to further reduce the risk of leakage of the mask array R and the substitution key K in the communication path with the key generating device 3A.

[0154] 3. Second embodiment In the first embodiment, distributed processing was performed on all elements of each dimension included in the feature data X. In contrast, in the second embodiment, a mode in which the registration information generating device 1A performs distributed processing on elements of some dimensions included in the feature data X to generate registration data T will be described. Note that the hardware configurations and functional configurations of the registration information generating device 1A, the matching information generating device 2A, the key generating device 3A, and the matching device 4A are the same as those of the first embodiment. In the second embodiment, differences from the first embodiment will be described, and the same configurations and modes as those of the first embodiment will not be described.

[0155] <3.1. Process flow for generating registration information> Fig. 16 is a flowchart showing the flow of a process for generating registration information in the registration information generating device 1A according to this embodiment. Fig. 17 is an explanatory diagram modeling the process for generating registration information in the registration information generating device 1A according to this embodiment. The flow of cancelable conversion in the registration information generating device 1A will be described with reference to Figs. 16 and 17.

[0156] The biometric information of the user acquired by the registration information generating device 1A is input to the controller 100A via the network I / F 101A. In the explanation of Fig. 16 and Fig. 17, it is assumed that the face image information of the user A is input to the registration information generating device 1A. In this embodiment, the key generating device 3A transmits the data to the registration information generating device 1A.

number

[0157] First, in step S41, the feature extraction unit 110A generates feature data X, which is an n-dimensional vector, from the facial image information of the user A ((Equation 2-1), (see the top row of FIG. 17 ). Note that the registration information generating device 1A may acquire the feature data X via the sensor 19 or the I / F 15.

[0158] Next, in step S42, the distributed processing unit 120A performs distributed processing on some elements of the n-dimensional feature amount data X with reference to the index information. Here, the index information corresponds to information indicating that the element of the dimension identified by the data S is a subset included in the first to n-th dimensions of the n-dimensional data. For the i-th element of the n-dimensional feature amount data X, the distributed processing unit 120A performs distributed processing on some elements of the n-dimensional feature amount data X with reference to the index information. Here, the index information corresponds to information indicating that the element of the dimension identified by the data S is a subset included in the first to n-th dimensions of the n-dimensional data.

number

number

number

[0159] In the following explanation, the l1-th element x in the n-dimensional feature data X is l1 , , l k Element x of dimension lk The data X is a partitioned vector sd´ This may be referred to as “partially distributed data”. Note that Fig. 17 shows a model diagram of the flow of processing for n-dimensional feature amount data X when the first-dimensional element and the second-dimensional element of the n-dimensional feature amount data X are included in the element corresponding to data S.

[0160] (formula 4 In the partially distributed data shown in -2), the elements of the first dimension to the kth dimension contained in the n-dimensional feature amount data X are distributed. Therefore, the partially distributed data is data of a vector of n+k dimensions.

[0161] The distributed processing of this embodiment corresponds to a process in which the distributed processing unit 120A distributes each of elements of some dimensions of the feature data X into a plurality of random numbers. Here, the distributed processing unit 120A determines the elements of the dimension of the feature data X to be distributed based on the index information. The distributed processing unit 120A divides the elements of the dimension of the feature data X identified by the index information, among the n-dimensional elements included in the feature data X, into a plurality of random numbers.

[0162] In this way, the index information corresponds to information indicating elements to be distributed by the distribution processing unit 120A among the elements included in the feature data X. In this embodiment, dividing the first-dimension element x1 of the feature data X into multiple elements (for example, random numbers x1[1] and x1[2] in (Equation 4-2)) that cannot be restored to the first-dimension element x1 of the feature data X by itself may be referred to as "secret sharing".

[0163] In (Equation 4-2), the distributed processing unit 120A calculates l k Element x of dimension lkBy secretly sharing the lk [1] and a random number x lk [2]), the number of divisions of the element x of the n-dimensional feature data x is not limited to two. For example, the distributed processing unit 120A generates the 11-th element x of the n-dimensional feature data x. l1 Let the number of divisions be j, (l, j are any positive integers).

number

[0164] Next, in step S23, the mask processing unit 140A applies a mask array

number

number

[0165] Next, in step S44, the substitution processing unit 130A performs a substitution process to rearrange the elements of each dimension of the positive / negative converted partial shared data by a substitution function F using a substitution key K. The substitution key K is an example of information that defines how to rearrange the elements included in the target information, and corresponds to an encryption key that is randomly generated by the key generation device 3A. As described above, the substitution key K is information generated by the key generation device 3A, and is information transmitted from the key generation device 3A to the registration information generation device 1A in order to generate registration information. By performing the substitution process, the substitution processing unit 130A rearranges the positive / negative converted partial shared data as follows:

number

[0166] For example, the elements of each dimension of the partial variance data after positive / negative conversion of the registered data T (i, j are any positive integers) are expressed as follows: 6 )

[0167] The registered data T may be generated by omitting the masking process using the mask array R. Furthermore, the registered data T may be generated by switching the order of the masking process and the replacement process.

[0168] The registration information generating device 1A transmits the thus generated registration data T to the DB 5A in step S45. The DB 5A stores the received registration data T. Note that in step S45, the registration information generating device 1A may transmit the registration data T related to the biometric information of the user A to the DB 5A together with an identifier that can identify the data as information related to the user A.

[0169] In this way, the registration information generating device 1A performs a mask process and a replacement process after secretly sharing some elements included in the feature data X, thereby generating the registration data T. Therefore, if the replacement key K is not leaked, it is difficult to determine which element of the feature data X corresponds to an element included in the registration data T. In addition, since the registration data T is generated after secretly sharing the feature data X, even if both the mask array R and the replacement key K are leaked, it is difficult to distinguish that the registration data newly generated using the leaked mask array R and the replacement key K and the leaked registration data T are derived from the same biometric feature data. As described above, the registration information generating device 1A of this embodiment can further reduce the risk of information leakage.

[0170] <3.2. Process flow for generating matching information> Next, a process flow when the matching information generating device 2A performs cancelable conversion of input information in this embodiment will be described. After storing the registration information in DB5A, the matching system 1000A needs to use biometric information of the same living body as the registration information when performing matching to determine whether or not the biometric authentication can be performed. The matching information generating device 2A acquires the biometric information of the user via the sensor 19 or I / F 15, and performs cancelable conversion based on the acquired biometric information. Information generated by cancelable conversion of the biometric information by the matching information generating device 2A is an example of matching information to be matched against the registration information stored in DB5A.

[0171] Fig. 18 is a flowchart showing the flow of processing for generating matching information in the matching information generation device 2A according to this embodiment. Fig. 19 is an explanatory diagram modeling the processing for generating matching information in the matching information generation device 2A according to this embodiment. The flow of cancelable conversion in the matching information generation device 2A will be described with reference to Figs. 18 and 19.

[0172] The biometric information of the user acquired by the matching information generation device 2A is input to the controller 200A via the network I / F 201A. In the explanation of Figs. 18 and 19, it is assumed that face image information of the user A is input to the matching information generation device 2A. In this embodiment, the key generation device 3A transmits the data to the matching information generation device 2A.

number

[0173] First, in step S51, the feature extraction unit 210A generates feature data Y, which is an n-dimensional vector, from the facial image information of the user A ((Equation 3-1), see the top row of FIG. 19). Note that the matching information generation device 2A may acquire the feature data Y via the sensor 19 or the I / F 15.

[0174] Next, in step S52, the duplication processing unit 220A refers to the index information and performs duplication processing on some elements of the n-dimensional feature amount data Y. Here, the index information corresponds to information indicating that the element of the dimension identified by the data S is a subset included in the first to n-th dimensions of the n-dimensional data. For the i-th element of the n-dimensional feature amount data Y, the duplication processing unit 220A performs the following.

number

number

number

[0175] FIG. 19 shows a model diagram of the flow of processing for n-dimensional feature amount data Y in the case where the first-dimensional elements and the second-dimensional elements of the n-dimensional feature amount data Y are included in the elements corresponding to data S. The partial duplication data shown in (Equation 5-2) is obtained by duplication of the first-dimensional to k-dimensional elements included in the n-dimensional feature amount data Y. Therefore, Reproduction The posterior data is data of an n+k dimensional vector.

[0176] In the duplication process of this embodiment, the duplication processing unit 220A duplications elements of some dimensions of the feature amount data Y as follows: Reproduction Here, the duplication processing unit 220A determines the element of the dimension of the feature amount data Y to be duplicated based on the index information. The duplication processing unit 220A duplicates each of the elements of the dimension of the feature amount data Y identified by the index information among the n-dimensional elements included in the feature amount data Y.

[0177] In this way, the index information corresponds to information indicating which of the elements included in the feature amount data Y are to be copied by the copy processor 220A.

[0178] In addition, in the registration information generating device 1A, when some elements of the feature amount data X are divided into a division number n1, the duplication processing unit 220A performs duplication processing to multiply some elements of the feature amount data Y by n2. Here, n1=n2, and both n1 and n2 are integers equal to or greater than 2. In the following explanation, the n-dimensional feature amount data Y generated by duplication processing is n+k dimensional vector of data Y cp´ This may be referred to as "partially replicated data."

[0179] In this embodiment, distribution processing and replication processing are performed so that the number of divisions of the feature amount data X in the registration information generating device 1A is the same as the number of copies of the feature amount data Y in the matching information generating device 2A.

[0180] Next, in step S53, the mask processing unit 240A performs mask processing on the elements of each dimension of the partially copied data using the mask array R ((Equation 4-4), see the third row from the top in FIG. 19). As described above, the mask array R is information generated by the key generation device 3A. The key generation device 3A generates the mask array R, which is an n+k-dimensional vector, and transmits it to the matching information generation device 2A. In this embodiment, the key generation device 3A transmits the same mask array R to the registration information generation device 1A and the matching information generation device 2A.

[0181] The positive and negative signs of the elements of each dimension of the partial duplication data can be converted by a mask process of multiplying the elements of each dimension of the partial duplication data by the mask array R. In the following description, the partial duplication data after the mask process of multiplying the elements of each dimension by the mask array R may be referred to as the "positive / negative converted partial duplication data."

[0182] Next, in step S54, the substitution processing unit 230A performs a substitution process to rearrange elements of each dimension of the positive / negative converted partial duplicated data by a substitution function F using a substitution key K. The substitution key K is an example of information that defines how to rearrange elements included in the target information, and corresponds to an encryption key randomly generated by the key generating device 3A. As described above, the substitution key K is information generated by the key generating device 3A, and is information transmitted from the key generating device 3A to the matching information generating device 2A in order to generate matching information. In this embodiment, the key generating device 3A transmits the same substitution key K to the registration information generating device 1A and the matching information generating device 2A.

[0183] The substitution processing unit 230A performs substitution processing to convert the positive / negative converted partial duplicated data into

number

[0184] The same substitution key K is used in the registration information generating device 1A and the matching information generating device 2A. That is, in the substitution process performed by the matching information generating device 2A, the elements of each dimension included in the partial copy data after positive / negative conversion are rearranged in the same order as in the substitution process in the registration information generating device 1A. That is, the matching data T' (i, j are any positive integers) corresponds to data in which the elements of each dimension included in the partial copy data after positive / negative conversion are rearranged as shown in (Equation 3-3), for example.

[0185] The matching data T' may be generated by omitting the masking process using the mask array R. Furthermore, the matching data T' may be generated by switching the order of the masking process and the replacement process.

[0186] Next, in step S55, the network I / F 201A transmits the matching data T' to the matching device 4A. At this time, the network I / F 201A may transmit an identifier (user identifier) ​​that can be identified as information related to user A together with the matching data T' to the matching device 4A.

[0187] In this way, the matching information generation device 2A generates matching data T' by duplicating the elements included in the feature data Y and then performing masking and replacement processes. Therefore, if the replacement key K is not leaked, it is difficult to determine which element of the feature data Y corresponds to an element included in the matching data T'. As described above, the matching information generation device 2A of this embodiment can further reduce the risk of information leakage.

[0188] 3.3. Calculation of Similarity in the Second Embodiment In this embodiment, a matching device 4A matches biometric information input to a registration information generating device 1A with biometric information input to a matching information generating device 2A in the same manner as in the first embodiment (see FIG. 13).

[0189] In step S31, the network I / F 401A acquires the enrollment data T from the DB 5A. The DB 5A stores the enrollment data T related to the biometric information of the user A received from the enrollment information generating device 1A in step S45 together with an identifier that can identify the information as information related to the user A. The matching device 4A acquires, from the DB 5A, the enrollment data T related to the biometric information of the user corresponding to the user identifier received from the matching information generating device 2A in step S55, that is, the user A.

[0190] In step S32, the collation device 4A of The similarity calculation unit 410A calculates the inner product of the registered data T and the matching data T' as a normalized correlation between the registered data T and the matching data T'. Here, the inner product of the registered data T and the matching data T', each of which is an n+k-dimensional vector, is

number

number

[0191] In this embodiment as well, the similarity calculation section 410A calculates the inner product of the registered data T and the matching data T′ as the similarity between the feature amount data X and the feature amount data Y.

[0192] Next, in step S33, the matching processing unit 420A performs a matching process to match the biometric information related to the feature data X with the biometric information related to the feature data Y based on the similarity between the feature data X and the feature data Y calculated in step S32.

[0193] The verification device 4A may authenticate the user identified by the verification information based on the result of the verification process in step S33.

[0194] In addition, for the mask array R used in the registration information generating device 1A, the matching information generating device 2A uses a mask array

number

number

number

[0195] In this way, in the calculation process using the enrollment data T generated by the enrollment information generating device 1A according to this embodiment and the matching data T' generated by the matching information generating device 2A, it is possible to calculate the similarity by calculating the inner product once, so that it is possible to expect a high throughput similar to that of conventional cancelable biometrics. In addition, since the inner product of the enrollment data T and the matching data T' is equal to the inner product of the feature data X and the feature data Y, it is possible to maintain the same matching accuracy as that of conventional cancelable biometrics. Therefore, the enrollment information generating device 1A and the matching information generating device 2A according to this embodiment can realize a cancelable conversion that can simultaneously suppress deterioration of the matching accuracy and reduce the risk of information leakage.

[0196] In this embodiment, as in the modified example of the first embodiment, the elements provided in the key generating device 3A may be provided in the registration information generating device 1A and the matching information generating device 2A (see Figs. 14 and 15). In this way, the registration information generating device 1A and the matching information generating device 2A can perform the cancelable conversion without communicating with the key generating device 3A. This makes it possible to further reduce the risk of leakage of the mask array R and the substitution key K in the communication path with the key generating device 3A.

[0197] <4. Third embodiment> Next, a third embodiment of the present invention will be described. In the first and second embodiments, a distribution process is performed on at least a part of the elements of each dimension included in the feature data X. In contrast, in the third embodiment, a configuration is described in which a feature extracted from a user's biometric information is subjected to a distribution process and a duplication process, and matching is performed based on the generated information. Note that in the third embodiment, the same components as those in the first and second embodiments are given the same reference numerals, and duplicated descriptions are omitted.

[0198] <4.1. Overview of the operation of the matching system> Fig. 20 is a diagram showing an example of an operation form of a verification system 1000B according to the third embodiment of the present invention. As shown in Fig. 20, the verification system 1000B is configured by connecting a registration information generating device 1B, a matching information generating device 2B, a key generating device 3B, a verification device 4B, and a database (hereinafter sometimes referred to as "DB") 5B via a network 6.

[0199] The registration information generating device 1B, the matching information generating device 2B, the key generating device 3B, the matching device 4B, and the DB5B have the same hardware configurations as the registration information generating device 1A, the matching information generating device 2A, the key generating device 3A, the matching device 4A, and the DB5A according to the first and second embodiments, respectively (see FIG. 2). In the description of the third embodiment, the description of the hardware configurations of the registration information generating device 1B, the matching information generating device 2B, the key generating device 3B, the matching device 4B, and the DB5B will be omitted.

[0200] The registration information generating device 1B is an information processing device such as a computer or a server in which a program is installed that generates registration information to be registered in the verification system 1000B from a user's biometric information. The registration information generating device 1B is realized by, for example, a portable information processing terminal such as a smartphone, an ATM, a PC connected to a sensor that detects biometric information, or the like. The process of generating registration information in the registration information generating device 1B will be described in detail later.

[0201] The matching information generating device 2B is an information processing device such as a computer or a server in which a program is installed that generates matching information from a user's biometric information for matching with the registered information registered in the matching system 1000B. The matching information generating device 2B is realized by, for example, a portable information processing terminal such as a smartphone, an ATM, a PC connected to a sensor that detects biometric information, or the like. The process of generating matching information in the matching information generating device 2B will be described in detail later.

[0202] The key generating device 3B is an information processing device that generates a private key used when converting a user's biometric information into registration information or matching information. The key generating device 3B generates, for example, a random number, a mask array, and a substitution key as the private key. The key generating device 3B transmits the generated private key to the registration information generating device 1B and the matching information generating device 2B via the network 6.

[0203] The verification device 4B is an information processing device that collates the verification information with the registration information. For example, in a biometric authentication technique, the verification device 4B collates the feature amount of the biometric information of the user stored as the registration information with the feature amount of the biometric information of the user inputted for authentication. The verification system 1000B may authenticate the user identified by the verification information based on the result of the collation between the verification information and the registration information.

[0204] DB5B is a storage medium that stores the registration information generated by the registration information generating device 1B, and the verification device 4B collates the registration information stored in DB5B with the verification information. Although the verification device 4B and DB5B are shown as independent elements in Fig. 20, DB5B may be implemented in a storage medium such as an HDD included in the verification device 4B.

[0205] In addition, while FIG. 20 shows an example in which the registration information generating device 1B and the matching information generating device 2B are connected to the network 6, they do not necessarily need to be connected to the network 6. For example, the registration information generating device 1B or the matching information generating device 2B may be connected to the key generating device 3B, the matching device 4B, and the DB5B by USB or the like. Furthermore, the registration information generating device 1B may be connected to the DB5B when transmitting the registration information to the DB5B. Furthermore, the matching information generating device 2B may be connected to the matching device 4B when transmitting the matching information. Furthermore, the key generating device 3B may be connected to the registration information generating device 1B or the matching information generating device 2B when transmitting the private key.

[0206] In addition, in the hardware configuration shown in Figure 2, the CPU 11 of registration information generating device 1B performs calculations according to a program stored in ROM 12 of registration information generating device 1B or a program loaded from storage medium 14 of registration information generating device 1B to RAM 13 of registration information generating device 1A, thereby forming a software control unit of registration information generating device 1B.

[0207] A functional block that realizes the functions of the controller 100B (see FIG. 21) of the registration information generating device 1B according to this embodiment is configured by combining the software control unit configured as described above with hardware.

[0208] Furthermore, in the hardware configuration as shown in FIG. 2, a software control unit of the matching information generating device 2B is configured by the CPU 11 of the matching information generating device 2B performing calculations in accordance with a program stored in the ROM 12 of the matching information generating device 2B or a program loaded from the storage medium 14 of the matching information generating device 2B to the RAM 13 of the matching information generating device 2B.

[0209] A functional block that realizes the functions of the controller 200B (see FIG. 22) of the matching information generation device 2B according to this embodiment is configured by combining the software control unit configured as described above with hardware.

[0210] In addition, in the hardware configuration as shown in FIG. 2, a software control unit of the key generation device 3B is configured by the CPU 11 of the key generation device 3B performing calculations in accordance with a program stored in the ROM 12 of the key generation device 3B or a program loaded from the storage medium 14 of the key generation device 3B to the RAM 13 of the key generation device 3B.

[0211] A functional block that realizes the functions of the controller 300B (see FIG. 23) of the key generation device 3B according to this embodiment is configured by combining the software control unit configured as described above with hardware.

[0212] In addition, in the hardware configuration as shown in FIG. 2, the CPU 11 of the matching device 4B performs calculations according to a program stored in the ROM 12 of the matching device 4B or a program loaded from the storage medium 14 of the matching device 4B to the RAM 13 of the matching device 4B, thereby forming a software control unit of the matching device 4B.

[0213] A functional block that realizes the functions of the controller 400B (see FIG. 24) of the verification device 4B according to this embodiment is configured by combining the software control unit configured as described above with hardware.

[0214] <4.2. Functional configuration of the registration information generating device> Next, the functional configuration of the registration information generating device 1B will be described with reference to Fig. 21. Fig. 21 is a functional block diagram illustrating the functional configuration of the registration information generating device 1B. As shown in Fig. 21, the registration information generating device 1B includes a controller 100B and a network I / F 101B which is an interface for the controller 100B to exchange information with other devices via a network 6.

[0215] The controller 100B generates registration information used in the verification system 1000B from information acquired via the network I / F 101B. The registration information is one of the pieces of information used for verification in the verification system 1000B. The registration information generated by the registration information generating device 1B is stored in the DB5B. The controller 100B is realized by installing a dedicated software program in an information processing device such as the registration information generating device 1B. The controller 100B includes a feature extraction unit 110B, an intermediate information generating unit 120B, a replacement processing unit 130B, and a mask processing unit 140B.

[0216] During authentication, the feature extraction unit 110B extracts features from the biometric information input to the controller 100B via the sensor 19 or I / F 15 of the registration information generating device 1B. In this embodiment, feature data X, which is an n-dimensional vector, is extracted from biometric information such as a fingerprint, vein, iris, and face image.

[0217] The intermediate information generating unit 120B generates intermediate information from the input information by performing distribution processing and replication processing. Details of the distribution processing and replication processing in this embodiment will be described later.

[0218] The substitution processing unit 130B performs substitution processing for substituting the arrangement of elements of the input information based on a private key such as the substitution key K. Details of the substitution processing in this embodiment will be described later.

[0219] The mask processing unit 140B performs a mask process to convert the elements of the input information using a mask array. The details of the mask process in this embodiment will be described later.

[0220] <4.3. Functional configuration of the matching information generating device> Next, the functional configuration of the matching information generating device 2B will be described with reference to Fig. 22. Fig. 22 is a functional block diagram illustrating the functional configuration of the matching information generating device 2B. As shown in Fig. 22, the matching information generating device 2B includes a controller 200B and a network I / F 201B which is an interface for the controller 200B to exchange information with other devices via a network 6.

[0221] The controller 200B generates matching information used in the matching system 1000B from information acquired via the network I / F 201B. The matching information is one of the pieces of information used for matching in the matching system 1000B. The matching information generated by the matching information generating device 2B is transmitted to the matching device 4B. The controller 200B is realized by installing a dedicated software program in an information processing device such as the matching information generating device 2B. The controller 200B includes a feature extraction unit 210B, an intermediate information generating unit 220B, a replacement processing unit 230B, and a mask processing unit 240B.

[0222] During authentication, the feature extraction unit 210B extracts features from biometric information input to the controller 200B via the sensor 19 or I / F 15 of the matching information generation device 2B. In this embodiment, feature data, which is an n-dimensional vector, is extracted from biometric information such as a fingerprint, vein, iris, and face image. Y Extract.

[0223] The intermediate information generating unit 220B generates intermediate information from the input information by performing distribution processing and replication processing. Details of the distribution processing and replication processing in this embodiment will be described later.

[0224] The substitution processing unit 230B performs substitution processing to substitute the arrangement of the elements of the input information based on a private key such as the substitution key K. Details of the substitution processing in this embodiment will be described later.

[0225] The mask processing unit 240B performs a mask process to convert the elements of the input information using a mask array. The details of the mask process in this embodiment will be described later.

[0226] <4.4. Functional configuration of the key generation device> Next, the functional configuration of the key generation device 3B will be described with reference to Fig. 23. Fig. 23 is a functional block diagram illustrating the functional configuration of the key generation device 3B. As shown in Fig. 23, the key generation device 3B includes a controller 300B and a network I / F 301B which is an interface for the controller 300B to exchange information with other devices via a network 6.

[0227] The controller 300B generates a private key used in the verification system 1000B from information acquired via the network I / F 301B. The controller 300B is realized by installing a dedicated software program in an information processing device such as the key generating device 3B. The controller 300B includes an index information generating unit 310B, a substitution key generating unit 320B, and a mask array generating unit 330B.

[0228] The index information generating unit 310B generates index information for identifying what kind of processing is to be performed on each element included in the information. As described above, the registration information generating device 1B and the matching information generating device 2B each perform distributed processing and duplication processing on the elements included in the information. The index information generating unit 310B generates, for example, information indicating the elements to be subjected to distributed processing as the index information. The index information generating unit 310B may generate index information including information indicating the elements to be subjected to distributed processing and information indicating the elements to be subjected to duplication processing. The network I / F 301B transmits the index information generated by the index information generating unit 310B to the registration information generating device 1B and the matching information generating device 2B.

[0229] The replacement key generating unit 320B generates a private key such as the replacement key K used by the registration information generating device 1B and the matching information generating device 2B in the replacement process. The network I / F 301B transmits the private key such as the replacement key K generated by the replacement key generating unit 320B to the registration information generating device 1B and the matching information generating device 2B.

[0230] The mask sequence generation unit 330B generates a mask sequence used in the mask process by the registration information generation device 1B and the matching information generation device 2B. The network I / F 301B transmits the mask sequence generated by the mask sequence generation unit 330B to the registration information generation device 1B and the matching information generation device 2B.

[0231] <4.5. Functional configuration of the verification device> Next, the functional configuration of the collation device 4B will be described with reference to Fig. 24. Fig. 24 is a functional block diagram illustrating the functional configuration of the collation device 4A. As shown in Fig. 24, the collation device 4B includes a controller 400B and a network I / F 401B that is an interface for the controller 400B to exchange information with other devices via the network 6.

[0232] The controller 400B acquires information input to the matching system 1000B via the network I / F 401B, and performs matching based on the acquired information. The controller 400B is realized by installing a dedicated software program in an information processing device such as the matching device 4B. The controller 400B includes a similarity calculation unit 410B and a matching processing unit 420B.

[0233] The similarity calculation unit 410B calculates a similarity indicating how similar each piece of information is. In this embodiment, the similarity calculation unit 410B calculates the similarity based on the normalized correlation between the registered information and the collation information. The details of the process performed by the similarity calculation unit 410B will be described later.

[0234] The matching processing unit 420B performs a matching process to determine whether or not a plurality of pieces of information match each other based on the similarity calculated by the similarity calculation unit 410B. For example, when the similarity calculation unit 410B calculates the similarity between the registered information and the matching information, the matching processing unit 420B performs the matching process to determine whether or not the matching information matches the registered information based on the similarity calculated by the similarity calculation unit 410B.

[0235] With the configuration described above, the verification system 1000B calculates the similarity between the registration information generated by the registration information generating device 1B and the verification information generated by the verification information generating device 2B. Then, based on the calculated similarity, it is determined whether the registration information and the verification information match each other. Note that information on the determination result indicating whether the verification information in the verification process matches the registration information may be used as information for determining whether authentication is possible in an authentication system having the verification system 1000B.

[0236] <4.6. Process flow for generating registration information> Next, a process flow when the registration information generating device 1B performs cancelable conversion of input information in this embodiment will be described. In the verification system 1000B, when performing verification to determine whether or not biometric authentication is possible, first, registration information, which is one of the pieces of information used for verification, is generated from the biometric information of the user input to the verification system 1000B. The registration information generating device 1B acquires the biometric information of the user input to the verification system 1000B via the sensor 19 or I / F 15 (see FIG. 2), and generates registration information based on the acquired biometric information.

[0237] Fig. 25 is a flowchart showing the flow of a process for generating registration information in the registration information generating device 1B according to this embodiment. Fig. 26 is an explanatory diagram modeling the process for generating registration information in the registration information generating device 1B according to this embodiment. The flow of cancelable conversion in the registration information generating device 1B will be described with reference to Figs. 25 and 26.

[0238] The biometric information of the user acquired by the registration information generating device 1B is input to the controller 100B via the network I / F 101B. In the explanation of Figs. 25 and 26, it is assumed that the facial image information of the user B is input to the registration information generating device 1B. In this embodiment, the key generating device 3B transmits the data to the registration information generating device 1B.

number

[0239] First, in step S61, the feature extraction unit 110B generates feature data X, which is an n-dimensional vector, from the facial image information of the user B ((Equation 2-1), see the top row of FIG. 26). Note that the registration information generating device 1B may acquire the feature data X via the sensor 19 or the I / F 15.

[0240] Next, in step S62, the intermediate information generating unit 120B performs distribution processing on some elements of the n-dimensional feature amount data X with reference to the index information. The index information corresponds to information indicating that the element of the dimension identified by the data S is a subset included in the first to n-th dimensions of the n-dimensional data, as shown in (Equation 6-1). The intermediate information generating unit 120B performs distribution processing on some elements of the n-dimensional feature amount data X with reference to the index information.

number

number

number

[0241] FIG. 26 shows a model diagram of the flow of processing for n-dimensional feature data X in a case where the first and second dimensional elements of the n-dimensional feature data X correspond to data S, and the (n-1)th and nth dimensional elements of the n-dimensional feature data X do not correspond to data S.

[0242] The registration intermediate data shown in (Formula 6-2) is obtained by distributing the elements of the first dimension to the kth dimension contained in the feature data X, which is an n-dimensional vector, and duplicating the elements of dimensions other than the first dimension to the kth dimension. Therefore, the registration intermediate data corresponds to data in which either a distribution process or a duplication process has been performed on the elements of each dimension of the feature data X, which is an n-dimensional vector. Therefore, the registration intermediate data is data of a 2n-dimensional vector.

[0243] The distribution process of this embodiment corresponds to a process in which the intermediate information generation unit 120B distributes each of elements of some dimensions of the feature data X into a plurality of random numbers. Here, the intermediate information generation unit 120B determines the dimensional elements of the feature data X to be distributed based on the index information. The intermediate information generation unit 120B divides each of the first-dimensional elements to the k-th-dimensional elements of the feature data X identified by the index information, out of the n-dimensional elements included in the feature data X, into a plurality of random numbers.

[0244] Moreover, the duplication process of this embodiment corresponds to a process in which the intermediate information generating unit 120B duplicates elements of some dimensions of the feature amount data X. Furthermore, in the distributed process for the feature amount data X, when the elements of the first dimension to the kth dimension of the feature amount data X are divided into a division number n1, the intermediate information generating unit 120B multiplies the elements of dimensions other than the elements of the first dimension to the kth dimension of the feature amount data X by n2 (n1=n2, n1 and n2 are both positive integers equal to or greater than 2) in the duplication process. In other words, the intermediate information generating unit 120B performs the distribution process and the duplication process so that the division number n1 of the elements of the feature amount data X in the distributed process is the same as the duplication number n2 of the elements of the feature amount data X.

[0245] In this way, the index information corresponds to information indicating, among the elements included in the feature amount data X, the elements to be distributed in the distribution process and the elements to be copied in the copy process. i The i-th element x of the feature data X alone i Dividing a random number into multiple elements that cannot be restored to its original state is sometimes called “secret sharing.” For example, in (Equation 6-2), the first-dimension element x1 of the feature data X is secretly shared into random numbers x1[1] and x1[2].

[0246] In (Equation 6-2), the intermediate information generating unit 120B secretly shares the first-dimensional element x1 of the n-dimensional feature data X to generate two random numbers (random numbers x1[1] and x1[2]). l1 The number of divisions is not limited to 2. For example, the intermediate information generating unit 120B divides the n-dimensional feature amount data X into the 11-th element x l1 Let the number of divisions be j (l, j are any positive integers), and the element x of the l1th dimension of the n-dimensional feature data X l1 may be secretly shared (see Equation 6-3).

number

[0247] Next, in step S63, the mask processing unit 140B is a mask array for each dimension of the intermediate data on the registration side.

number

number

[0248] Next, in step S64, the substitution processing unit 130B performs a substitution process to rearrange the elements of each dimension of the positive / negative converted registration side intermediate data by a substitution function F using a substitution key K. The substitution key K is an example of information that defines how to rearrange the elements included in the target information, and corresponds to an encryption key that is randomly generated by the key generation device 3B. As described above, the substitution key K is information generated by the key generation device 3B, and is information transmitted from the key generation device 3B to the registration information generation device 1B. By performing the substitution process, the substitution processing unit 130B rearranges the positive / negative converted registration side intermediate data into

number

[0249] The registered data T (i, j are any positive integers) corresponds to data in which the elements of each dimension included in the registered intermediate data after positive / negative conversion are rearranged as shown in (Equation 6-5), for example.

number

number

[0250] The registered data T may be generated by omitting the masking process using the mask array R. Furthermore, the registered data T may be generated by switching the order of the masking process and the replacement process.

[0251] The registration information generating device 1B transmits the thus generated registration data T to the DB 5B in step S65. The DB 5B stores the received registration data T. Note that in step S65, the registration information generating device 1B may transmit the registration data T related to the biometric information of the user B to the DB 5B together with an identifier that can identify the registration data T as information related to the user B.

[0252] In this way, the registration information generating device 1B performs secret sharing of the elements included in the feature data X, and then performs a mask process and a replacement process to generate the registration data T. Therefore, if the replacement key K is not leaked, it is difficult to determine which element of the feature data X corresponds to the element included in the registration data T. Furthermore, since the feature data X is secretly shared and copied before the registration data T is generated, even if both the mask array R and the replacement key K are leaked, it is difficult to distinguish that the registration data newly generated using the leaked mask array R and the replacement key K and the leaked registration data T are derived from the same biometric feature data. As described above, the registration information generating device 1B of this embodiment can further reduce the risk of information leakage.

[0253] <4.7. Process flow for generating matching information> Next, a process flow when the matching information generating device 2B performs cancelable conversion of input information in this embodiment will be described. After storing the registration information in DB5B, the matching system 1000B needs to use biometric information of the same living body as the registration information when performing matching to determine whether or not the biometric authentication can be performed. The matching information generating device 2B acquires the biometric information of the user via the sensor 19 or I / F 15, and performs cancelable conversion based on the acquired biometric information. Information generated by cancelable conversion of the biometric information by the matching information generating device 2B is an example of matching information to be matched against the registration information stored in DB5B.

[0254] Fig. 27 is a flowchart showing the flow of processing for generating matching information in the matching information generating device 2B according to this embodiment. Fig. 28 is an explanatory diagram modeling the processing for generating matching information in the matching information generating device 2B according to this embodiment. The flow of cancelable conversion in the matching information generating device 2B will be described with reference to Figs. 27 and 28.

[0255] The biometric information of the user acquired by the matching information generation device 2B is input to the controller 200B via the network I / F 201B. In the explanation of Figs. 27 and 28, it is assumed that face image information of user B is input to the matching information generation device 2B. In this embodiment, the key generation device 3B transmits the data to the matching information generation device 2B.

number

[0256] First, in step S71, feature extraction unit 210B generates feature data Y, which is an n-dimensional vector, from facial image information of user B ((Equation 3-1), see the top row of FIG. 28). Note that matching information generation device 2B may acquire feature data Y via sensor 19 or I / F 15.

[0257] Next, in step S72, the intermediate information generating unit 220B refers to the index information to generate intermediate information from the n-dimensional feature amount data Y. The index information corresponds to information indicating that the element of the dimension identified by the data S is a subset included in the first to n-th dimensions of the n-dimensional data. In step S72, the intermediate information generating unit 220B generates intermediate information for the i-th element of the n-dimensional feature amount data Y by:

number

number

number

number

number

[0258] Note that Figure 28 shows a model diagram of the processing flow for n-dimensional feature data Y in the case where the first and second dimensional elements of the n-dimensional feature data Y are elements that correspond to data S, and the (n-1)th and nth dimensional elements of the n-dimensional feature data Y are elements that do not correspond to data S.

[0259] The matching-side intermediate data shown in (Formula 7-3) is obtained by duplicating the elements of the first to kth dimensions contained in the feature data Y, which is an n-dimensional vector, and distributing the elements of dimensions other than the first to kth dimensions. Therefore, the matching-side intermediate data corresponds to data in which either a distribution process or a duplication process has been performed on the elements of each dimension of the feature data Y, which is an n-dimensional vector. Therefore, the matching-side intermediate data is data of a 2n-dimensional vector.

[0260] The distribution process of this embodiment corresponds to a process in which the intermediate information generation unit 220B distributes elements of some dimensions of the feature amount data Y into a plurality of random numbers. Here, the intermediate information generation unit 220B determines the elements of the dimensions of the feature amount data Y to be distributed based on the index information. The intermediate information generation unit 220B divides, among the n-dimensional elements included in the feature amount data Y, the elements of the dimensions other than the first to k-th dimensions of the feature amount data X identified by the index information into a plurality of random numbers.

[0261] Moreover, the duplication process of this embodiment corresponds to a process in which the intermediate information generating unit 220B duplicates elements of some dimensions of the feature amount data Y. Furthermore, in the distributed process for the feature amount data Y, when elements of dimensions other than the first to k-th dimensions of the feature amount data Y are divided by the division number n1, the intermediate information generating unit 220B multiplies the elements of the first to k-th dimensions of the feature amount data Y by n2 (n1=n2, n1 and n2 are both positive integers equal to or greater than 2) in the duplication process. In other words, the intermediate information generating unit 220B performs the distribution process and the duplication process so that the division number n1 of the elements of the feature amount data Y in the distributed process is the same as the duplication number n2 of the elements of the feature amount data Y.

[0262] The intermediate information generating unit 220B performs distribution processing in the registration information generating device 1B so that the number of divisions when the elements of the feature amount data X are divided is equal to the number of divisions n1 when the elements of the feature amount data Y are divided. Also, the intermediate information generating unit 220B performs distribution processing in the registration information generating device 1B so that the number of copies when the elements of the feature amount data X are copied is equal to the number of copies n2 when the elements of the feature amount data Y are copied. Reproduction Perform processing.

[0263] In this way, the index information corresponds to information indicating, among the elements included in the feature amount data Y, the elements to be distributed in the distribution process and the elements to be copied in the copy process. i The i-th element y of the feature data Y is i For example, in (Equation 7-3), the n-th element y of the feature data Y is n Let y be a random number. n [1] and a random number y n [2] It is being secretly distributed.

[0264] In (Equation 7-3), the intermediate information generating unit 220B calculates the n-th element y n By secretly sharing the n [1] and a random number y n [2]) is shown, but the element y of n-dimensional feature data Y is n The number of divisions is not limited to 2. For example, the intermediate information generating unit 220B divides the n-th element y n Let the number of divisions be j (j is any positive integer), and the n-th element y of the n-dimensional feature data Y n may be secretly shared (see Equation 7-4).

number

[0265] Next, in step S73, the mask processing unit 240B performs mask processing on the elements of each dimension of the matching side intermediate data using the mask array R (see formula 6-4, third row from the top in FIG. 28). As described above, the mask array R is information generated by the key generation device 3B. The key generation device 3B generates the mask array R, which is a 2n-dimensional vector, and transmits it to the matching information generation device 2B. Note that the mask processing of multiplying the elements of each dimension of the matching side intermediate data by the mask array R can convert the positive and negative of the elements of each dimension of the matching side intermediate data. Note that in the following description, the matching side intermediate data after the mask processing of multiplying the elements of each dimension by the mask array R may be referred to as "matching side intermediate data after positive / negative conversion."

[0266] Next, in step S74, the substitution processing unit 230B performs substitution processing to rearrange the elements of each dimension of the positive / negative converted matching side intermediate data by a substitution function F using a substitution key K. The substitution key K is an example of information that defines how to rearrange the elements included in the target information, and corresponds to an encryption key that is randomly generated by the key generation device 3B. As described above, the substitution key K is information generated by the key generation device 3B, and is information transmitted from the key generation device 3B to the matching information generation device 2B. By performing substitution processing, the substitution processing unit 230B rearranges the positive / negative converted matching side intermediate data into

number

[0267] The registration information generating device 1B and the matching information generating device 2B use the same substitution key K. That is, in the substitution process performed by the matching information generating device 2B, the elements of each dimension included in the positive / negative converted matching side intermediate data are rearranged in the same order as in the substitution process in the registration information generating device 1B.

[0268] Therefore, the matching data T' (i, j are any positive integers) corresponds to data in which the elements of each dimension contained in the matching side intermediate data after positive / negative conversion are rearranged as shown in (equation 7-5) by the replacement process in the registration information generation device 1B (see (equation 6-5) and (equation 6-6)).

number

[0269] Also, the matching data T' (i', j' are any positive integers) corresponds to data in which the elements of each dimension included in the matching side intermediate data after positive / negative conversion are rearranged as shown in (Equation 7-6).

number

[0270] The matching data T' may be generated by omitting the masking process using the mask array R. Furthermore, the matching data T' may be generated by switching the order of the masking process and the replacement process.

[0271] The matching information generating device 2B transmits the matching data T' thus generated to the matching device 4B in step S75. The matching device 4B stores the received matching data T'. At this time, the matching information generating device 2B may transmit an identifier (user identifier) ​​that can identify the matching data T' as information related to user B to the matching device 4B together with the matching data T'.

[0272] In this way, the matching information generating device 2B performs secret sharing of the elements included in the feature data Y, and then performs masking and substitution processing to generate the matching data T'. Therefore, if the substitution key K is not leaked, it is difficult to determine which element of the feature data Y corresponds to the element included in the matching data T'. In addition, since the matching data T' is generated after secret sharing and copying of the feature data Y, even if both the mask array R and the substitution key K are leaked, it is difficult to distinguish that the matching data newly generated using the leaked mask array R and substitution key K and the leaked matching data T' are derived from the same biometric feature data. As described above, the matching information generating device 2B of this embodiment can further reduce the risk of information leakage.

[0273] <4.8. Process flow for calculating similarity> Next, a process flow in the matching device 4B in this embodiment will be described with reference to Fig. 29. Fig. 29 is a flowchart showing a process flow in the matching device 4B for matching the biometric information input to the registration information generating device 1B with the biometric information input to the matching information generating device 2B.

[0274] In step S81, the network I / F 401B acquires the enrollment data T from the DB 5B. The DB 5B stores the enrollment data T related to the biometric information of the user B received from the enrollment information generating device 1B in step S65 together with an identifier that can identify the information as related to the user B. The matching device 4B acquires, from the DB 5B, the enrollment data T related to the biometric information of the user corresponding to the user identifier received from the matching information generating device 2B in step S75, that is, the user B.

[0275] Subsequently, in step S82, similarity calculation section 410B calculates the similarity between feature amount data X and feature amount data Y.

[0276] In step S82, the similarity calculation unit 410B calculates the inner product of the enrollment data T and the matching data T' as a normalized correlation between the enrollment data T and the matching data T'. Here, the inner product of the enrollment data T and the matching data T', each of which is a 2n-dimensional vector,

number

number

[0277] Furthermore, for any positive integers i=i', j=j', from (equation 2-2), (equation 6-6), and (equation 7-6), (equation 8-1) can be expressed as follows:

number

number

[0278] Next, in step S83, the matching processing unit 420B performs a matching process to match the biometric information related to the feature data X with the biometric information related to the feature data Y based on the similarity between the feature data X and the feature data Y calculated in step S82.

[0279] The verification device 4B may authenticate the user identified by the verification information based on the result of the verification process in step S83.

[0280] In addition, for the mask array R used in the registration information generating device 1B, the mask array

number

number

number

[0281] In this way, in the calculation process using the enrollment data T generated by the enrollment information generating device 1B according to this embodiment and the matching data T' generated by the matching information generating device 2B, it is possible to calculate the similarity by calculating the inner product once, so that it is possible to expect a high throughput similar to that of conventional cancelable biometrics. In addition, since the inner product of the enrollment data T and the matching data T' is equal to the inner product of the feature data X and the feature data Y, it is possible to maintain the same matching accuracy as that of conventional cancelable biometrics. Therefore, the enrollment information generating device 1B and the matching information generating device 2B according to this embodiment can realize a cancelable conversion that can simultaneously suppress the deterioration of the matching accuracy and reduce the risk of information leakage.

[0282] <4.9. Modifications of the registration information generating device and the matching information generating device> The registration information generating device 1B and the matching information generating device 2B may each include an element that generates the mask array R and the substitution key K. FIG. 30 is a functional block diagram showing the functional configuration of the registration information generating device 1B according to a modified example of the third embodiment. 31 13 is a functional block diagram showing a functional configuration of a matching information generating device 2B according to a modified example of the third embodiment. FIG.

[0283] As shown in FIG. 30, the registration information generating device 1B includes a feature extraction unit 110B, an intermediate information generating unit 120B In addition to the substitution processing unit 130B and the mask processing unit 140B (see FIG. 21), the system further includes a substitution key generation unit 150B, an index information generation unit 160B and a mask sequence generation unit 170B.

[0284] The substitution key generating unit 150B generates a private key such as a substitution key K used by the registration information generating device 1B in the substitution process. The index information generating unit 160B generates index information, which is information for identifying what process is to be performed on each element included in the information. The mask array generating unit 170B generates a mask array used by the registration information generating device 1B in the mask process.

[0285] As shown in FIG. 31, the matching information generation device 2B includes a feature extraction unit 210B, an intermediate information generation unit 220B, a substitution processing unit 230B, and a mask processing unit 240B (see FIG. 22), as well as a substitution key generation unit 250B, an index information generation unit 260B, and a mask array generation unit 270B.

[0286] The substitution key generation unit 250B generates a private key such as a substitution key K used by the matching information generation device 2B in the substitution process. The index information generation unit 260B generates index information, which is information for identifying what process is to be performed on each element included in the information. The mask array generation unit 270B generates a mask array used by the matching information generation device 2B in the mask process.

[0287] According to this modification, the registration information generating device 1B and the matching information generating device 2B can perform the cancelable conversion without communicating with the key generating device 3B. This makes it possible to further reduce the risk of leakage of the mask array R and the substitution key K in the communication path with the key generating device 3B.

[0288] <5. Fourth embodiment> In the third embodiment, the registration information is generated by performing distribution processing or duplication processing on the elements of each dimension included in the feature data X. In contrast, in the fourth embodiment, a mode in which the registration information generating device 1B performs at least one of distribution processing or duplication processing on the elements included in the feature data X to generate registration information will be described. In this embodiment, when a character with a hat is indicated, "(hat)" is written after the character. The hardware configuration and functional configuration of the registration information generating device 1B, the matching information generating device 2B, the key generating device 3B, and the matching device 4B are the same as those of the third embodiment. In the fourth embodiment, the differences from the third embodiment will be described, and the same configuration and aspects as those of the third embodiment will be omitted.

[0289] <5.1. Process flow for generating registration information> Fig. 32 is a flowchart showing the flow of a process for generating registration information in the registration information generating device 1B according to this embodiment. Fig. 33 is an explanatory diagram modeling the process for generating registration information in the registration information generating device 1B according to this embodiment. The flow of cancelable conversion in the registration information generating device 1B will be described with reference to Figs. 32 and 33.

[0290] The biometric information of the user acquired by the registration information generating device 1B is input to the controller 100B via the network I / F 101B. In the explanation of Fig. 32 and Fig. 33, it is assumed that the facial image information of the user B is input to the registration information generating device 1B. In this embodiment, the key generating device 3B transmits the data to the registration information generating device 1B.

number

number

number

[0291] First, in step S91, the feature extraction unit 110B generates feature data X, which is an n-dimensional vector, from the facial image information of user B ((Equation 2-1), see the top row of FIG. 33). Note that the registration information generating device 1B may acquire the feature data X via the sensor 19 or the I / F 15. The feature data X is an example of registration side input information in this embodiment.

[0292] Next, in step S92, the intermediate information generating unit 120B performs distribution processing on some elements of the n-dimensional feature amount data X with reference to the index information. The index information corresponds to information indicating that the elements of the dimension indicated by the data S are a subset of the n-dimensional data. For the i-th element of the n-dimensional feature amount data X, the intermediate information generating unit 120B performs distribution processing on some elements of the n-dimensional feature amount data X with reference to the index information.

number

number

number

[0293] FIG. 33 shows a model diagram of the flow of processing for n-dimensional feature data X in the case where the first and second dimensional elements of the n-dimensional feature data X are elements corresponding to data S, the third dimensional element of the n-dimensional feature data X is an element that does not correspond to either data S or data S(hat), and the (n-1)th and nth dimensional elements of the n-dimensional feature data X are elements that correspond to data S(hat).

[0294] According to (Equation 9-1), in step S92, the intermediate information generating unit 120B divides each of the first-dimension element x1 and the second-dimension element x2 included in the n-dimensional feature amount data X into a plurality of random numbers. 9 In 2, the intermediate information generating unit 120B calculates the n-1th element x included in the n-dimensional feature amount data X. n-1 and the n-th element x n As a result, the first-dimensional element x1 and the second-dimensional element x2 contained in the n-dimensional feature data X are divided, and the n-1-dimensional element x n-1 and the n-th element x n Data where is a replicated vector

number

[0295] The distribution process of this embodiment corresponds to a process in which the intermediate information generation unit 120B distributes each of elements of some dimensions of the feature amount data X into a plurality of random numbers. Here, the intermediate information generation unit 120B determines the elements of the dimensions of the feature amount data X to be distributed based on the index information. The intermediate information generation unit 120B divides the elements of the dimension corresponding to the data S, among the n-dimensional elements included in the feature amount data X, into a plurality of random numbers.

[0296] Moreover, the duplication process of this embodiment corresponds to a process in which the intermediate information generating unit 120B duplicates elements of some dimensions of the feature amount data X. Furthermore, in the distributed process for the feature amount data X, when elements of a dimension corresponding to data S among n-dimensional elements included in the feature amount data X are divided by a division number n1, the intermediate information generating unit 120B multiplies elements of a dimension corresponding to data S (hat) among n-dimensional elements included in the feature amount data X by n2 (n1=n2, n1 and n2 are both positive integers equal to or greater than 2) in the duplication process. In other words, the intermediate information generating unit 120B performs the distribution process and the duplication process so that the division number n1 of the elements of the feature amount data X in the distributed process is the same as the duplication number n2 of the elements of the feature amount data X.

[0297] In this way, the index information corresponds to information indicating elements to be distributed in the distribution process and elements to be copied in the copy process among the elements included in the feature data X. In this embodiment, for example, among the n-dimensional elements included in the feature data X, an element of a dimension corresponding to the data S is an example of an element corresponding to the first set, and an element of a dimension corresponding to the data S (hat) is an example of an element corresponding to the second set.

[0298] In this embodiment, the i-th element x of the feature amount data X i The i-th element x of the feature data X alone iDividing the feature data X into multiple elements that cannot be restored to its original state is sometimes called “secret sharing.” For example, in (Equation 9-4), the first-dimension element x1 of the feature data X is secretly shared into the random numbers x1[1] and x1[2].

[0299] In (Equation 9-4), the intermediate information generating unit 120B secretly shares the first-dimension element x1 of the n-dimensional feature data X to generate two random numbers (random number x1[1] and random number x1[2]), but the number of divisions is not limited to n1=2. For example, the intermediate information generating unit 120B secretly shares the first-dimension element x1 of the n-dimensional feature data X to generate two random numbers (random number x1[1] and random number x1[2]). l1 Let the number of divisions be p (p is a positive integer greater than or equal to 2), and the n-dimensional feature data X's l1-th element x l1 may be secretly shared (see equation 9-5).

number

[0300] Next, in step S93, the mask processing unit 140B applies a mask array

number

number

[0301] Next, in step S94, the substitution processing unit 130B performs a substitution process to rearrange the elements of each dimension of the registration side intermediate data after positive / negative conversion by a substitution function F using a substitution key K. The substitution process performed by the substitution processing unit 130B is an example of the registration side substitution process of this embodiment. The substitution key K is an example of information that defines how to rearrange the elements included in the target information, and corresponds to an encryption key that is randomly generated by the key generation device 3B. As described above, the substitution key K is information generated by the key generation device 3B, and is information transmitted from the key generation device 3B to the registration information generation device 1B. By performing the substitution process, the substitution processing unit 130B rearranges the registration side intermediate data after positive / negative conversion into

number

[0302] The registered data T (i, j are any positive integers) corresponds to data in which the elements of each dimension included in the registered intermediate data after positive / negative conversion are rearranged as shown in (Equation 9-7), for example.

number

number

[0303] The registered data T may be generated by omitting the masking process using the mask array R. Furthermore, the registered data T may be generated by switching the order of the masking process and the replacement process.

[0304] In step S95, the registration information generating device 1B transmits the thus generated registration data T to the DB 5B. The DB 5B stores the received registration data T. In step S95, the registration information generating device 1B may transmit the registration data T related to the biometric information of the user B to the DB 5B together with an identifier that can identify the registration data T as information related to the user B.

[0305] In this way, the registration information generating device 1B performs secret sharing of the elements included in the feature data X, and then performs a mask process and a replacement process to generate the registration data T. Therefore, if the replacement key K is not leaked, it is difficult to determine which element of the feature data X corresponds to the element included in the registration data T. Furthermore, since the feature data X is secretly shared and copied before the registration data T is generated, even if both the mask array R and the replacement key K are leaked, it is difficult to distinguish that the registration data newly generated using the leaked mask array R and the replacement key K and the leaked registration data T are derived from the same biometric feature data. As described above, the registration information generating device 1B of this embodiment can further reduce the risk of information leakage.

[0306] <5.2. Process flow for generating matching information> Next, a process flow when the matching information generating device 2B performs cancelable conversion of input information in this embodiment will be described. After storing the registration information in DB5B, the matching system 1000B needs to use biometric information of the same living body as the registration information when performing matching to determine whether or not the biometric authentication can be performed. The matching information generating device 2B acquires the biometric information of the user via the sensor 19 or I / F 15, and performs cancelable conversion based on the acquired biometric information. Information generated by cancelable conversion of the biometric information by the matching information generating device 2B is an example of matching information to be matched against the registration information stored in DB5B.

[0307] Fig. 34 is a flowchart showing the flow of processing for generating matching information in the matching information generating device 2B according to this embodiment. Fig. 35 is an explanatory diagram modeling the processing for generating matching information in the matching information generating device 2B according to this embodiment. The flow of cancelable conversion in the matching information generating device 2B will be described with reference to Figs. 34 and 35.

[0308] The biometric information of the user acquired by the matching information generation device 2B is input to the controller 200B via the network I / F 201B. In the explanation of Fig. 34 and Fig. 35, it is assumed that facial image information of the user B is input to the matching information generation device 2B.

[0309] In this embodiment, the key generating device 3B transmits information indicating that at least one of the data S and S(hat) is a subset included in the n-dimensional data as index information to the matching information generating device 2B (see (Equation 9-1), (Equation 9-2), and (Equation 9-3)). The matching information generating device 2B generates matching information from the facial image information of user B by referring to the index information. For the sake of explanation, the explanation will be continued assuming that both of the data S and S(hat) are subsets included in the n-dimensional data. Note that, according to (Equation 9-3), elements included in both of the data S and S(hat) are not included in the n-dimensional data.

[0310] First, in step S101, feature extraction unit 210B generates feature data Y, which is an n-dimensional vector, from facial image information of user B ((Equation 3-1), see the top row of FIG. 35). Note that matching information generation device 2B may acquire feature data Y via sensor 19 or I / F 15. The feature data Y is an example of matching-side input information in this embodiment.

[0311] Next, in step S102, the intermediate information generating unit 220B refers to the index information to generate n-dimensional feature amount data. Y The index information corresponds to information indicating that the element of the dimension identified by the data S is a subset of the n-dimensional data. In step S102, the intermediate information generating unit 220B performs distributed processing on some of the elements of the n-dimensional feature amount data Y by

number

number

number

[0312] FIG. 35 shows a model diagram of the processing flow for n-dimensional feature data Y in the case where the first and second dimensional elements of the n-dimensional feature data Y are elements that correspond to data S, the third dimensional element of the n-dimensional feature data Y is an element that does not correspond to either data S or data S(hat), and the (n-1)th dimensional element and the nth dimensional element of the n-dimensional feature data Y are elements that correspond to data S(hat).

[0313] According to (Equation 9-1), in step S102, the intermediate information generating unit 220B copies the first-dimensional element y1 and the second-dimensional element y2 included in the n-dimensional feature amount data Y. Also, according to (Equation 9-2), in step S102, the intermediate information generating unit 220B copies the (n-1)th-dimensional element y n-1 and the n-th element y n are divided into multiple random numbers. As a result, the first-dimensional element y1 and the second-dimensional element y2 contained in the n-dimensional feature data Y are each duplicated, and the n-1th-dimensional element y n-1 and the n-th element y nData where is a partitioned vector

number

[0314] The distribution process of this embodiment corresponds to a process in which the intermediate information generating unit 220B distributes elements of some dimensions of the feature amount data Y into a plurality of random numbers. Y The intermediate information generating unit 220B divides each of the n-dimensional elements included in the feature amount data Y, that is, the elements of a dimension corresponding to the data S (hat), into a plurality of random numbers.

[0315] In the duplication process of this embodiment, the intermediate information generating unit 220B generates feature amount data Y This corresponds to a process of duplicating elements of some dimensions in the feature data Y. In addition, in the distributed processing for the feature data Y, when elements of a dimension corresponding to data S (hat) among n-dimensional elements included in the feature data Y are divided as a division number n1, the intermediate information generating unit 220B multiplies elements of a dimension corresponding to data S among n-dimensional elements included in the feature data Y by n2 (n1=n2, n1 and n2 are both positive integers equal to or greater than 2) in the duplication processing. In other words, the intermediate information generating unit 220B performs the distribution processing and duplication processing so that the division number n1 of the elements of the feature data Y in the distributed processing is the same as the duplication number n2 of the elements of the feature data Y.

[0316] In this way, the index information corresponds to information indicating elements to be distributed in the distribution process and elements to be copied in the copy process among the elements included in the feature data Y. In this embodiment, for example, among the n-dimensional elements included in the feature data Y, an element of a dimension corresponding to data S is an example of an element corresponding to the first set, and an element of a dimension corresponding to data S (hat) is an example of an element corresponding to the second set.

[0317] In this embodiment, the i-th element y of the feature amount data Y i The i-th element y of the feature data Y is i For example, in (Equation 10-1), the n-1th element y of the feature data Y is n-1 Let y be a random number. n-1 [1] and a random number y n-1 [2] It is being secretly distributed.

[0318] In (Equation 10-1), the intermediate information generating unit 220B calculates the n-1-th element y n-1 By secretly sharing the n-1 [1] and a random number y n-1 [2]) is shown, the number of divisions is not limited to n1=2. For example, the intermediate information generating unit 220B generates the j1-th element of the n-dimensional feature amount data y. y j1 Let the number of divisions be p (p is a positive integer greater than or equal to 2), and Y j1st dimension element of y j1 may be secretly shared (see equation 10-2).

number

[0319] Next, in step S103, the mask processing unit 240B Matching side Mask processing is performed on the elements of each dimension of the intermediate data using a mask array R ((Equation 9-6), see the third row from the top in FIG. 35). The mask processing performed by mask processing unit 240B is an example of matching-side mask processing in this embodiment. As described above, mask array R is information generated by key generation device 3B. Key generation device 3B generates mask array R, which is an n+k+m-dimensional vector, and transmits it to matching information generation device 2B. In this embodiment, key generation device 3B transmits the same mask array R to registration information generation device 1B and matching information generation device 2B. Each element included in mask array R is,

number

[0320] Next, in step S104, the substitution processing unit 230B performs a substitution process to rearrange elements of each dimension of the positive / negative converted matching side intermediate data by a substitution function F using a substitution key K. The substitution process performed by the substitution processing unit 230B is an example of the matching side substitution process of this embodiment. The substitution key K is an example of information that defines how to rearrange elements included in the target information, and corresponds to an encryption key that is randomly generated by the key generation device 3B. As described above, the substitution key K is information generated by the key generation device 3B, and is information transmitted from the key generation device 3B to the matching information generation device 2B. By performing the substitution process, the substitution processing unit 230B rearranges the positive / negative converted matching side intermediate data into

number

[0321] The registration information generating device 1B and the matching information generating device 2B use the same substitution key K. That is, in the substitution process performed by the matching information generating device 2B, the elements of each dimension included in the positive / negative converted matching side intermediate data are rearranged in the same order as in the substitution process in the registration information generating device 1B.

[0322] Therefore, the matching data T' (i, j are any positive integers) corresponds to data in which the elements of each dimension included in the matching side intermediate data after positive / negative conversion are rearranged as shown in (Equation 10-3).

number

[0323] Also, the matching data T' (i', j' are any positive integers) corresponds to data in which the elements of each dimension included in the matching side intermediate data after positive / negative conversion are rearranged as shown in (Equation 10-4).

number

[0324] The matching data T' may be generated by omitting the masking process using the mask array R. Furthermore, the matching data T' may be generated by switching the order of the masking process and the replacement process.

[0325] The matching information generating device 2B transmits the matching data T' thus generated to the matching device 4B in step S75. The matching device 4B stores the received matching data T'. At this time, the matching information generating device 2B may transmit an identifier (user identifier) ​​that can identify the matching data T' as information related to user B to the matching device 4B together with the matching data T'.

[0326] In this way, the matching information generating device 2B performs secret sharing of the elements included in the feature data Y, and then performs masking and substitution processing to generate the matching data T'. Therefore, if the substitution key K is not leaked, it is difficult to determine which element of the feature data Y corresponds to the element included in the matching data T'. In addition, since the matching data T' is generated after secret sharing and copying of the feature data Y, even if both the mask array R and the substitution key K are leaked, it is difficult to distinguish that the matching data newly generated using the leaked mask array R and substitution key K and the leaked matching data T' are derived from the same biometric feature data. As described above, the matching information generating device 2B of this embodiment can further reduce the risk of information leakage.

[0327] <5.3. Process flow for calculating similarity> In this embodiment, a matching device 4B matches biometric information input to a registration information generating device 1B with biometric information input to a matching information generating device 2B in the same manner as in the third embodiment (see FIG. 29).

[0328] In step S81, the network I / F 401B acquires the registration data T from the DB 5B. The DB 5B receives the user information T from the registration information generating device 1B in step S65. B The verification device 4B stores the registration data T related to the biometric information of the user A together with an identifier that can identify that the information is related to the user A. The verification device 4B stores the user corresponding to the user identifier received from the verification information generating device 2B in step S75, that is, the user B Registration data T relating to the biometric information of the person is obtained from DB5B.

[0329] In step S82, the similarity calculation unit 410B calculates the inner product of the registered data T and the matching data T' as a normalized correlation between the registered data T and the matching data T'. Here, the inner product of the registered data T and the matching data T', each of which is an n+k+m-dimensional vector,

number

number

[0330] Furthermore, for any positive integers i=i', j=j', from (equation 7-2), (equation 9-8), and (equation 10-4), (equation 8-3) can be expressed as follows:

number

[0331] In this embodiment as well, the similarity calculation section 410B calculates the inner product of the registered data T and the matching data T′ as the similarity between the feature amount data X and the feature amount data Y.

[0332] Next, in step S83, the matching processing unit 420B performs a matching process to match the biometric information related to the feature data X with the biometric information related to the feature data Y based on the similarity between the feature data X and the feature data Y calculated in step S82.

[0333] The verification device 4B may authenticate the user identified by the verification information based on the result of the verification process in step S83.

[0334] In addition, for the mask array R used in the registration information generating device 1B, the mask array

number

number

number

[0335] In this way, in the calculation process using the enrollment data T generated by the enrollment information generating device 1B according to this embodiment and the matching data T' generated by the matching information generating device 2B, it is possible to calculate the similarity by calculating the inner product once, so that it is possible to expect a high throughput similar to that of conventional cancelable biometrics. In addition, since the inner product of the enrollment data T and the matching data T' is equal to the inner product of the feature data X and the feature data Y, it is possible to maintain the same matching accuracy as that of conventional cancelable biometrics. Therefore, the enrollment information generating device 1B and the matching information generating device 2B according to this embodiment can realize a cancelable conversion that can simultaneously suppress the deterioration of the matching accuracy and reduce the risk of information leakage.

[0336] In this embodiment, similarly to the modified example of the third embodiment, the elements provided in the key generating device 3B may be provided in the registration information generating device 1B and the matching information generating device 2B (see Figs. 30 and 31). In this way, the registration information generating device 1B and the matching information generating device 2B can perform the cancelable conversion without communicating with the key generating device 3B. This makes it possible to further reduce the risk of leakage of the mask array R, the substitution key K, and the index information in the communication path with the key generating device 3B.

[0337] <6. Application to Euclidean distance> The similarity between feature data X and Y generated from biometric information can also be calculated by Euclidean distance in addition to normalized correlation. Fig. 36 is a model diagram of cancelable transformation for Euclidean distance as a second reference example of the present invention.

[0338] In FIG. 36, the feature quantity extracted from the biometric information is represented as n-dimensional vector feature quantity data.

number

[0339] Feature data X E When generating registration information for face recognition from E For each dimension of the elements, the mask array

number

[0340] Next, the randomly transformed feature data X E The order of elements in each dimension of is the permutation key

number

number

[0341] Next, when generating matching information for matching, the features extracted from the biometric information used for matching are converted into feature data, which is an n-dimensional vector.

number

number

[0342] And the registration data T E and matching data T´ E The similarity to the registered data T E and matching data T´ Et The Euclidean distance between

number

number

[0343] Then, for any positive integers i, j,

number

number

[0344] However, even in an authentication system using cancelable biometrics for Euclidean distance, the mask array R E Masking and substitution key K E Therefore, if an attacker finds data that is close to feature data X,

number

[0345] In addition, if an attacker finds data similar to feature data X,

number

number

[0346] In addition, the registration data T E If the registration data T E The same mask array R used to generate E and replacement key K E Using the above, the registration data T E New biometric feature data X of the same person related to E 2 to registration data T E Let's assume that you generate 2 and register it in the authentication system.

[0347] In this case, the registration data T E Feature data X relating to E And feature data XE Since the data is close to 2, the registered data T E 2, the leaked registration data T E This makes it easy to carry out so-called spoofing, which involves matching using the same information.

[0348] Furthermore, registration information T E In case that is leaked, in order to ensure the security of the entire authentication system, the mask array R E and replacement key K E It is necessary to update the registration data T E into a new mask array R E and replacement key K E must be updated to the registration information generated using

[0349] On the other hand, when calculating the similarity of features extracted from biometric information using Euclidean distance, the n-dimensional feature data X in (Equation 2-1) is expressed as follows, based on (Equation 11-1):

number

number

[0350] When the similarity of features extracted from biometric information is calculated using Euclidean distance, the inner product of transformed feature data X', which is a vector, and transformed feature data Y', which is a vector, is calculated using the Euclidean distance.

number

[0351] <7. Fifth embodiment> Next, a fifth embodiment of the present invention will be described with reference to Figures 38 to 40. The first and second embodiments described above are specific embodiments, but the fifth embodiment is a more generalized embodiment.

[0352] Fig. 38 is a block diagram illustrating a schematic configuration of a matching system 1000C according to a fifth embodiment of the present invention. As shown in Fig. 38, the matching system 1000C includes a registration information generating device 1C, a matching information generating device 2C, and a matching device 4C including a similarity calculation unit that calculates a similarity between registration side input information and matching side input information based on the registration information and the matching information.

[0353] 39 is a block diagram illustrating a schematic configuration of a registration information generating device 1C according to the fifth embodiment. The registration information generating device 1C includes a distribution processing unit 110C and a registration side replacement processing unit 120C.

[0354] The sharing processing unit 110C performs sharing processing to secretly share elements included in the input information, thereby generating secret sharing information from registration side input information input to the registration information generating device 1C for registration. The registration side substitution processing unit 120C performs registration side substitution processing to substitute the order of elements included in the input information, using a substitution key. The registration information generating device 1C generates registration information by substituting the order of elements included in the secret sharing information, using the substitution key.

[0355] 40 is a block diagram illustrating a schematic configuration of a matching information generation device 2C according to the fifth embodiment. The matching information generation device 2C includes a duplication processing unit 210C and a matching side replacement processing unit 220C.

[0356] The duplication processing unit 210C performs duplication processing to duplicate elements contained in the input information, thereby generating duplicate information from matching-side input information input to the matching information generation device 2C for matching. The matching-side substitution processing unit 220C performs matching-side substitution processing to substitute the order of elements contained in the input information, using a substitution key. The matching information generation device 2C generates matching information by substituting the order of elements contained in the duplicate information, using the substitution key.

[0357] -Relationship with the first and second embodiments As an example, the registration information generating device 1C according to the fifth embodiment may execute the operation of the registration information generating device 1A according to the first or second embodiment. Similarly, as an example, the matching information generating device 2C according to the fifth embodiment may execute the operation of the matching information generating device 2A according to the first or second embodiment. Similarly, as an example, the matching system 1000C according to the fifth embodiment may execute the operation of the matching system 1000A according to the first or second embodiment. In the above case, the description of the first embodiment or the second embodiment is also applicable to the fifth embodiment. Note that the fifth embodiment is not limited to the above examples.

[0358] 8. Sixth embodiment Next, a sixth embodiment of the present invention will be described with reference to Figures 41 to 43. The third and fourth embodiments described above are specific embodiments, but the sixth embodiment is a more generalized embodiment.

[0359] Fig. 41 is a block diagram illustrating a schematic configuration of a matching system 1000D according to a sixth embodiment of the present invention. As shown in Fig. 41, the matching system 1000D includes a registration information generating device 1D, a matching information generating device 2D, and a matching device 4D including a similarity calculation unit that calculates a similarity between registration side input information and matching side input information based on the registration information and the matching information.

[0360] 42 is a block diagram illustrating a schematic configuration of a registration information generating device 1D according to the sixth embodiment. The registration information generating device 1D includes a registration side intermediate information generating unit 110D and a registration side replacement processing unit 120D.

[0361] The registration side intermediate information generation unit 110D performs a registration side sharing process for secretly sharing elements corresponding to a first set of input information and a registration side copying process for copying elements corresponding to a second set of input information, based on the registration side input information input to the registration information generation device 1D for registration, thereby generating registration side intermediate information from the registration side input information. The registration side replacement processing unit 120D performs a registration side replacement process for replacing the order of elements included in the input information using a replacement key. The registration information generation device 1D generates registration information by replacing the order of elements included in the registration side intermediate information using the replacement key.

[0362] 43 is a block diagram illustrating a schematic configuration of a matching information generation device 2D according to the sixth embodiment. The matching information generation device 2D includes a matching-side intermediate information generation unit 210D and a matching-side replacement processing unit 220D.

[0363] The matching-side intermediate information generation unit 210D performs a matching-side duplication process for duplication of elements corresponding to a first set of input information and a matching-side sharing process for secretly sharing elements corresponding to a second set of input information, based on the matching-side input information input to the matching information generation device 2D for matching, thereby generating matching-side intermediate information from the matching-side input information. The matching-side substitution processing unit 220D performs a matching-side permutation process for permuting the order of elements included in the input information using a substitution key. The matching information generation device 2D generates matching information by permuting the order of elements included in the matching-side intermediate information using the substitution key.

[0364] -Relationship with the third and fourth embodiments As an example, the registration information generating device 1D according to the sixth embodiment may execute the operation of the registration information generating device 1B according to the third or fourth embodiment.D However, the verification system 1000D according to the sixth embodiment may execute the operation of the verification information generating device 2B according to the third or fourth embodiment. B In the above cases, the explanation of the third embodiment or the fourth embodiment can also be applied to the sixth embodiment. Note that the sixth embodiment is not limited to the above examples.

[0365] <9. Other embodiments> Although the embodiments of the present invention have been described above, the present invention is not limited to these embodiments. It will be understood by those skilled in the art that these embodiments are merely examples and that various modifications are possible without departing from the scope and spirit of the present invention.

[0366] For example, the steps in the processes described herein do not necessarily have to be performed chronologically according to the order depicted in the flowchart diagrams. For example, the steps in the processes may be performed in an order different from that depicted in the flowcharts, or may be performed in parallel. Also, some of the steps in the processes may be removed, and additional steps may be added to the processes.

[0367] Also, a device (e.g., one or more devices (or units) among the multiple devices (or units) constituting the registration information generating device, or a module for one of the multiple devices (or units)) including the components of the registration information generating device described in this specification (the distributed processing unit, the registration side intermediate information generating unit, and / or the registration side replacement processing unit) may be provided. A device (e.g., one or more devices (or units) among the multiple devices (or units) constituting the matching information generating device, or a module for one of the multiple devices (or units)) including the components of the matching information generating device described in this specification (the duplication processing unit, the matching side intermediate information generating unit, and / or the matching side replacement processing unit) may be provided. Also, a method including the processing of the components may be provided, or a program for causing a processor to execute the processing of the components may be provided. Also, a non-transitory computer readable medium having the program recorded thereon may be provided. Naturally, such a device, module, method, program, and non-transitory computer readable medium are also included in the present invention.

[0368] A part or all of the above-described embodiments may be described as, but is not limited to, the following supplementary notes.

[0369] (Appendix 1) A registration information generating device that generates registration information based on registration side input information input for registration, the registration side input information includes at least one of an element corresponding to a first set and an element corresponding to a second set different from the first set, a registration side intermediate information generation unit that generates registration side intermediate information from the registration side input information by performing a registration side sharing process of secretly sharing elements corresponding to the first set of input information and a registration side duplication process of duplication of elements corresponding to the second set of input information; a registration side replacement processing unit that performs a registration side replacement process for replacing the order of elements included in the input information by using a replacement key; The registration information is the registration side intermediate information is generated by replacing the order of elements included in the registration side intermediate information with the replacement key; Registration information generating device.

[0370] (Appendix 2) a registration mask processing unit that performs a registration mask process for converting elements included in input information using a mask array; The registration information is the registration intermediate information is generated by converting elements included in the registration intermediate information using the mask array and replacing the order of the elements included in the registration intermediate information using the replacement key; 2. A registration information generating device as described in appendix 1.

[0371] (Appendix 3) a registration-side mask sequence generation unit that generates the mask sequence, 3. A registration information generating device as described in appendix 2.

[0372] (Appendix 4) The registration side intermediate information generation unit In the registration side distributed processing, each of the elements corresponding to the first set of the input information is divided into n1 pieces of shared information (n is an integer greater than or equal to 2), and in the registration side replication processing, the elements corresponding to the second set of the input information are multiplied by n2 (n is an integer greater than or equal to 2) to generate the registration side intermediate information; The n1 pieces of shared information are different values, a sum of the n1 pieces of shared information is equal to an element corresponding to the first set of the input information before being divided into the n1 pieces of shared information. 4. A registration information generating device according to any one of claims 1 to 3.

[0373] (Appendix 5) a registration-side replacement key generation unit that generates the replacement key, 5. A registration information generating device according to any one of appendix 1 to 4.

[0374] (Appendix 6) a registration index information generating unit that generates registration index information indicating the elements of the first set and the elements of the second set; 6. A registration information generating device according to any one of appendix 1 to 5.

[0375] (Appendix 7) A registration information generating device that generates registration information based on registration side input information input for registration, a distribution processing unit that performs a secret sharing process on elements included in input information to generate secret sharing information from the registration side input information; a registration side replacement processing unit that performs a registration side replacement process for replacing the order of elements included in the input information by using a replacement key; The registration information is The secret sharing information is generated by replacing the order of elements included in the secret sharing information with the replacement key. Registration information generating device.

[0376] (Appendix 8) a registration mask processing unit that performs a registration mask process for converting elements included in input information using a mask array; The registration information is The secret sharing information is generated by converting elements included in the secret sharing information using the mask array, and by replacing the order of the elements included in the secret sharing information using the replacement key. 8. A registration information generating device as described in appendix 7.

[0377] (Appendix 9) a registration-side mask sequence generation unit that generates the mask sequence, A registration information generating device as described in appendix 8.

[0378] (Appendix 10) The distributed processing unit includes: In the sharing process, each element included in the input information is divided into n1 pieces of shared information (n≧2) to generate the secret shared information; The n1 pieces of shared information are each a different value, a sum of the n1 pieces of shared information is equal to an element included in the input information before being divided into the n1 pieces of shared information; 10. A registration information generating device according to any one of appendix 7 to 9.

[0379] (Appendix 11) a registration-side replacement key generation unit that generates the replacement key, 11. A registration information generating device according to any one of appendix 7 to 10.

[0380] (Appendix 12) A matching information generating device that generates matching information based on matching side input information input for matching, comprising: the matching input information includes at least one of an element corresponding to a first set and an element corresponding to a second set different from the first set, a matching-side intermediate information generating unit that generates matching-side intermediate information from the matching-side input information by performing a matching-side duplication process of duplication of elements corresponding to the first set of input information and a matching-side sharing process of secret sharing of elements corresponding to the second set of input information; a matching-side permutation processing unit that performs a matching-side permutation process for permuting an order of elements included in input information by using a substitution key, The collation information is the verification intermediate information is generated by replacing the order of elements included in the verification intermediate information by using the replacement key; Matching information generating device.

[0381] (Appendix 13) a matching-side mask processing unit that performs matching-side mask array processing for converting elements included in input information using a mask array; The collation information is the verification-side intermediate information is generated by converting elements included in the verification-side intermediate information using the mask sequence and replacing the order of the elements included in the verification-side intermediate information using the replacement key; 13. The matching information generating device according to claim 12.

[0382] (Appendix 14) a matching mask sequence generation unit that generates the mask sequence, 14. The matching information generating device according to claim 13.

[0383] (Appendix 15) The matching side intermediate information generation unit in the matching-side replication process, an element corresponding to the first set of the input information is multiplied by n1 (n is an integer greater than or equal to 2), and in the matching-side distribution process, each element corresponding to the second set of the input information is divided into n2 pieces of shared information (n is an integer greater than or equal to 2) to generate the matching-side intermediate information; the n1 multiple is the same as the number of divisions when dividing each of the elements corresponding to the first set of input information into shared information in a registration information generation device that generates registration information to be matched with the matching information, The n2 pieces of shared information are each a different value, a sum of the n2 pieces of shared information is equal to an element corresponding to the second set of the input information before being divided into the n2 pieces of shared information; The n2 is the same as the number of copies of each element of the second set of input information in the registration information generation device. 15. The matching information generating device according to any one of appendix 12 to 14.

[0384] (Appendix 16) a matching-side replacement key generation unit that generates the replacement key, 16. The matching information generating device according to any one of appendix 12 to 15.

[0385] (Appendix 17) a matching index information generating unit that generates matching index information indicating the elements of the first set and the elements of the second set; 17. The matching information generating device according to any one of appendix 12 to 16.

[0386] (Appendix 18) A matching information generating device that generates matching information based on matching side input information input for matching, comprising: a duplication processing unit that performs duplication processing for duplication of elements included in input information to generate duplicate information from the matching input information; a matching-side permutation processing unit that performs a matching-side permutation process for permuting an order of elements included in input information by using a substitution key, The collation information is The duplicate information is generated by replacing an order of elements included in the duplicate information with the replacement key. Matching information generating device.

[0387] (Appendix 19) a matching-side mask array processing unit that performs matching-side mask processing for converting elements included in input information using a mask array; The collation information is The duplicate information is generated by converting elements included in the duplicate information using the mask array and replacing the order of the elements included in the duplicate information using the replacement key. 19. The matching information generating device of claim 18.

[0388] (Appendix 20) a matching mask sequence generation unit that generates the mask sequence, 20. The matching information generating device of claim 19.

[0389] (Appendix 21) The duplication processing unit includes: In the duplication process, an element included in the input information is multiplied by n2 (n2 is an integer equal to or greater than 2) to generate the duplication information; The n2 times is the same as the division number n1 (n1 is an integer greater than or equal to 2) when dividing each of the elements included in the input information into shared information in a registration information generation device that generates registration information to be matched with the matching information. 21. The matching information generating device according to any one of appendix 18 to 20.

[0390] (Appendix 22) a matching-side replacement key generation unit that generates the replacement key, 22. The matching information generating device according to any one of appendix 18 to 21.

[0391] (Appendix 23) A registration information generating device according to any one of claims 1 to 6, A matching information generating device according to any one of claims 12 to 17, a matching device having a similarity calculation unit that calculates a similarity between the registration side input information and the matching side input information based on the registration information and the matching information; A matching system comprising:

[0392] (Appendix 24) The collation system comprises: A replacement key generation unit that generates the replacement key; a mask sequence generation unit that generates a transmission mask sequence to be transmitted to the registration information generation device and the matching information generation device; and an index information generating unit that generates index information indicating the elements of the first set and the elements of the second set. 23. A matching system as described in Appendix 23.

[0393] (Appendix 25) A registration information generating device according to any one of claims 7 to 11, A matching information generating device according to any one of claims 18 to 22, a matching device having a similarity calculation unit that calculates a similarity between the registration side input information and the matching side input information based on the registration information and the matching information; A matching system comprising:

[0394] (Appendix 26) The collation system comprises: A replacement key generation unit that generates the replacement key; a mask sequence generation unit that generates a transmission mask sequence to be transmitted to the registration information generation device and the matching information generation device, A matching system as described in Appendix 25.

[0395] (Appendix 27) A method for generating registration information based on registration side input information input for registration, the registration information including at least one of an element corresponding to a first set and an element corresponding to a second set different from the first set, the method comprising: generating registration side intermediate information from the registration side input information by performing a registration side sharing process for secretly sharing elements corresponding to the first set of input information and a registration side copying process for copying elements corresponding to the second set of input information; performing a registration side replacement process for replacing the order of elements included in the input information by using a replacement key; generating the registration information by replacing the order of elements included in the registration side intermediate information by using the replacement key; How registration information is generated.

[0396] (Appendix 28) A registration information generation method for generating registration information based on registration side input information input for registration, comprising the steps of: generating secret sharing information from the registration side input information by performing a sharing process for secretly sharing elements included in the input information; performing a registration side replacement process for replacing the order of elements included in the input information by using a replacement key; generating the registration information by replacing the order of elements included in the secret sharing information with the replacement key; How registration information is generated.

[0397] (Appendix 29) A matching information generating method for generating matching information based on matching side input information input for matching, the matching information including at least one of an element corresponding to a first set and an element corresponding to a second set different from the first set, the matching information comprising: generating matching-side intermediate information from the matching-side input information by performing a matching-side duplication process for duplication of elements corresponding to the first set of input information and a matching-side sharing process for secretly sharing elements corresponding to the second set of input information; performing a matching side permutation process for permuting the order of elements included in the input information by using a permutation key; generating the matching information by replacing the order of elements included in the matching intermediate information by using the replacement key; How matching information is generated.

[0398] (Appendix 30) A matching information generating method for generating matching information based on matching side input information input for matching, comprising the steps of: generating duplicate information from the input information on the verification side by performing a duplication process of duplication of elements included in the input information; performing a matching side permutation process for permuting the order of elements included in the input information by using a permutation key; generating the verification information by replacing the order of elements included in the duplicate information with the replacement key; How matching information is generated.

[0399] (Appendix 31) A similarity calculation method for a verification system including a registration information generation device that generates registration information based on registration side input information input for registration, a verification information generation device that generates verification information based on verification side input information input for verification, and a verification device that verifies the registration side input information and the verification side input information, comprising: the registration-side input information and the matching-side input information each include at least one of an element corresponding to a first set and an element corresponding to a second set different from the first set, The registration information generating device, generating registration side intermediate information from the registration side input information by performing a registration side sharing process for secretly sharing elements corresponding to the first set of input information and a registration side copying process for copying elements corresponding to the second set of input information; performing a registration side replacement process for replacing the order of elements included in the input information by using a replacement key; generating the registration information by replacing the order of elements included in the registration side intermediate information with the replacement key; The matching information generating device, generating matching-side intermediate information from the matching-side input information by performing a matching-side duplication process for duplication of elements corresponding to the first set of input information and a matching-side sharing process for secretly sharing elements corresponding to the second set of input information; performing a matching-side permutation process for permuting the order of elements included in the input information by using the permutation key; generating the verification information by replacing an order of elements included in the verification side intermediate information with the replacement key; The collation device, calculating a similarity between the registration side input information and the matching side input information based on the registration information and the matching information; Similarity calculation method.

[0400] (Appendix 32) A similarity calculation method for a verification system including a registration information generation device that generates registration information based on registration side input information input for registration, a verification information generation device that generates verification information based on verification side input information input for verification, and a verification device that verifies the registration side input information and the verification side input information, comprising: The registration information generating device, generating shared information from the registration side input information by performing a sharing process for secretly sharing elements included in the input information; performing a registration side replacement process for replacing the order of elements included in the input information by using a replacement key; generating the registration information by replacing an order of elements included in the shared information by using the replacement key; The matching information generating device, generating duplicate information from the input information on the verification side by performing a duplication process of duplication of elements included in the input information; performing a matching-side permutation process for permuting the order of elements included in the input information by using the permutation key; generating the verification information by replacing an order of elements included in the duplicate information with the replacement key; The collation device, calculating a similarity between the registration side input information and the matching side input information based on the registration information and the matching information; Similarity calculation method.

[0401] (Appendix 33) A registration information generation program that generates registration information based on registration side input information input for registration, the registration information including at least one of an element corresponding to a first set and an element corresponding to a second set different from the first set, generating registration side intermediate information from the registration side input information by performing a registration side sharing process for secretly sharing elements corresponding to the first set of input information and a registration side copying process for copying elements corresponding to the second set of input information; performing a registration side replacement process for replacing the order of elements included in the input information by using a replacement key; generating the registration information by replacing the order of elements included in the registration side intermediate information by using the replacement key; Registration information generator.

[0402] (Appendix 34) A registration information generation program for generating registration information based on registration side input information input for registration, generating secret sharing information from the registration side input information by performing a sharing process for secretly sharing elements included in the input information; performing a registration side replacement process for replacing the order of elements included in the input information by using a replacement key; generating the registration information by replacing the order of elements included in the secret sharing information with the replacement key; Registration information generator.

[0403] (Appendix 35) A matching information generating method for generating matching information based on matching side input information input for matching, the matching information including at least one of an element corresponding to a first set and an element corresponding to a second set different from the first set, the matching information comprising: generating matching-side intermediate information from the matching-side input information by performing a matching-side duplication process for duplication of elements corresponding to the first set of input information and a matching-side sharing process for secretly sharing elements corresponding to the second set of input information; performing a matching side permutation process for permuting the order of elements included in the input information using a permutation key; generating the matching information by replacing the order of elements included in the matching intermediate information by using the replacement key; Matching information generator.

[0404] (Appendix 36) A verification information generating program for generating verification information based on verification side input information input for verification, generating duplicate information from the input information on the verification side by performing a duplication process of duplication of elements included in the input information; performing a matching side permutation process for permuting the order of elements included in the input information using a permutation key; generating the verification information by replacing the order of elements included in the duplicate information with the replacement key; Matching information generator. [Industrial Applicability]

[0405] The present invention provides a registration information generating device, a matching information generating device, a matching system, a registration information generating method, a matching information generating method, a similarity calculation method, a registration information generating program, and a matching information generating program that realizes a cancellable conversion that can simultaneously suppress deterioration of matching accuracy and reduce the risk of information leakage. [Explanation of symbols]

[0406] 1A, 1B, 1C, 1D Registration information generating device 2A, 2B, 2C, 2D verification information generation device 3A, 3B Key generation device 4A, 4B, 4C, 4D verification device 5A, 5B DB 110A, 110B Feature extraction unit 110C Distributed Processing Unit 110D Registration side intermediate information generation unit 120A Distributed Processing Unit 120B Intermediate information generation section 120C, 120D Registration side replacement processing unit 130A, 130B Replacement processing section 140A, 140B Mask processing section 210A, 210B Feature extraction unit 210C Replication processing unit 210D Matching side intermediate information generation unit 220A Replication Processing Unit 220B Intermediate information generation section 220C, 220D Matching side replacement processing unit 230A, 230B Replacement processing section 240A, 240B Mask processing section 310A, 310B Index information generating unit 320A, 320B Replacement key generation unit 330A, 330B Mask array generator 410A, 410B Similarity calculation unit 420A, 420B Matching processing section 1000A, 1000B, 1000C, 1000D Matching System

Claims

1. A registration information generating device that generates registration information based on registration side input information input for registration, the registration side input information includes at least one of an element corresponding to a first set and an element corresponding to a second set different from the first set, a registration side intermediate information generating means for generating registration side intermediate information from the registration side input information by performing a registration side sharing process for secretly sharing elements corresponding to the first set of input information and a registration side copying process for copying elements corresponding to the second set of input information; a registration side replacement processing means for performing a registration side replacement process for replacing the order of elements included in the input information by using a replacement key; The registration information is the registration side intermediate information is generated by replacing the order of elements included in the registration side intermediate information with the replacement key; Registration information generating device.

2. a registration mask processing means for performing registration mask processing for converting elements included in input information using a mask array; The registration information is the registration intermediate information is generated by converting elements included in the registration intermediate information using the mask array and replacing the order of the elements included in the registration intermediate information using the replacement key; The registration information generating device according to claim 1 .

3. a registration-side mask sequence generating means for generating the mask sequence, The registration information generating device according to claim 2 .

4. The registration side intermediate information generating means in the registration side distribution processing, dividing each of the elements corresponding to the first set of the input information into n1 pieces of shared information (n is an integer greater than or equal to 2); and in the registration side replication processing, generating the registration side intermediate information by multiplying the elements corresponding to the second set of the input information by n2 (n is an integer greater than or equal to 2); The n1 pieces of shared information are different values, a sum of the n1 pieces of shared information is equal to an element corresponding to the first set of the input information before being divided into the n1 pieces of shared information. The registration information generating device according to claim 1 .

5. A registration-side replacement key generating means for generating the replacement key is provided. The registration information generating device according to claim 1 .

6. a registration index information generating means for generating registration index information indicating the elements of the first set and the elements of the second set; The registration information generating device according to any one of claims 1 to 5.

7. A registration information generating device that generates registration information based on registration side input information input for registration, a sharing processing means for generating secret sharing information from the registration side input information by performing a sharing process for secretly sharing elements included in the input information; a registration side replacement processing means for performing a registration side replacement process for replacing the order of elements included in the input information by using a replacement key; The registration information is The secret sharing information is generated by replacing the order of elements included in the secret sharing information with the replacement key. Registration information generating device.

8. a registration mask processing means for performing registration mask processing for converting elements included in input information using a mask array; The registration information is The secret sharing information is generated by converting elements included in the secret sharing information using the mask array, and by replacing the order of the elements included in the secret sharing information using the replacement key. The registration information generating device according to claim 7.

9. a registration-side mask sequence generating means for generating the mask sequence, The registration information generating device according to claim 8.

10. The distributed processing means includes: In the sharing process, each element included in the input information is divided into n1 pieces of shared information (n≧2) to generate the secret shared information; The n1 pieces of shared information are each a different value, a sum of the n1 pieces of shared information is equal to an element included in the input information before being divided into the n1 pieces of shared information, The registration information generating device according to any one of claims 7 to 9.

11. A registration-side replacement key generating means for generating the replacement key is provided. The registration information generating device according to any one of claims 7 to 10.

12. A matching information generating device that generates matching information based on matching side input information input for matching, comprising: the matching input information includes at least one of an element corresponding to a first set and an element corresponding to a second set different from the first set, a verification-side intermediate information generating means for generating verification-side intermediate information from the verification-side input information by performing a verification-side duplication process for duplication of elements corresponding to the first set of input information and a verification-side sharing process for secretly sharing elements corresponding to the second set of input information; a matching-side permutation processing means for performing a matching-side permutation process for permuting the order of elements included in the input information by using a permutation key; The collation information is the verification intermediate information is generated by replacing the order of elements included in the verification intermediate information by using the replacement key; Matching information generating device.

13. a matching mask processing means for performing matching mask processing for converting elements included in input information using a mask array; The collation information is the verification-side intermediate information is generated by converting elements included in the verification-side intermediate information using the mask sequence and replacing the order of the elements included in the verification-side intermediate information using the replacement key; The matching information generating device according to claim 12.

14. a matching mask sequence generating means for generating the mask sequence, The matching information generating device according to claim 13.

15. The verification side intermediate information generating means in the matching-side duplication process, elements corresponding to the first set of the input information are multiplied by n1 (n is an integer greater than or equal to 2); in the matching-side distribution process, each element corresponding to the second set of the input information is divided into n2 pieces of shared information (n is an integer greater than or equal to 2), thereby generating the matching-side intermediate information; the n1 multiple is the same as the number of divisions when dividing each of the elements corresponding to the first set of input information into shared information in a registration information generation device that generates registration information to be matched with the matching information, The n2 pieces of shared information are each a different value, a sum of the n2 pieces of shared information is equal to an element corresponding to the second set of the input information before being divided into the n2 pieces of shared information, The n2 is the same as the number of copies of each element of the second set of input information in the registration information generation device. The matching information generating device according to any one of claims 12 to 14.

16. a matching-side replacement key generating means for generating the replacement key, The matching information generating device according to any one of claims 12 to 15.

17. a matching index information generating means for generating matching index information indicating the elements of the first set and the elements of the second set, The matching information generating device according to any one of claims 12 to 16.

18. A matching information generating device that generates matching information based on matching side input information input for matching, comprising: a duplication processing means for generating duplicate information from the input information on the verification side by performing a duplication process for duplication of elements included in the input information; a matching-side permutation processing means for performing a matching-side permutation process for permuting the order of elements included in the input information by using a permutation key; The collation information is The duplicate information is generated by replacing an order of elements included in the duplicate information with the replacement key. Matching information generating device.

19. a matching mask processing means for performing matching mask processing for converting elements included in input information using a mask array; The collation information is The duplicate information is generated by converting elements included in the duplicate information using the mask array and replacing the order of the elements included in the duplicate information using the replacement key. The matching information generating device according to claim 18.

20. a matching mask sequence generating means for generating the mask sequence, The matching information generating device according to claim 19.

21. The duplication processing means includes: In the duplication process, an element included in the input information is multiplied by n2 (n2 is an integer greater than or equal to 2) to generate the duplicated information; The n2 times is the same as the division number n1 (n1 is an integer greater than or equal to 2) when dividing each of the elements included in the input information into shared information in a registration information generation device that generates registration information to be matched with the matching information. The matching information generating device according to any one of claims 18 to 20.

22. a matching-side replacement key generating means for generating the replacement key, The matching information generating device according to any one of claims 18 to 21.

23. A registration information generating device according to any one of claims 1 to 6, A matching information generating device according to any one of claims 12 to 17, a matching device having a similarity calculation means for calculating a similarity between the registration side input information and the matching side input information based on the registration information and the matching information; A matching system comprising:

24. The collation system comprises: A replacement key generating means for generating the replacement key; a mask sequence generating means for generating a transmission mask sequence to be transmitted to the registration information generating device and the matching information generating device; and an index information generating means for generating index information indicating the elements of the first set and the elements of the second set.

24. The verification system of claim 23.

25. A registration information generating device according to any one of claims 7 to 11, A matching information generating device according to any one of claims 18 to 22, a matching device having a similarity calculation means for calculating a similarity between the registration side input information and the matching side input information based on the registration information and the matching information; A matching system comprising:

26. The collation system comprises: A replacement key generating means for generating the replacement key; a mask sequence generating means for generating a transmission mask sequence to be transmitted to the registration information generating device and the matching information generating device, 26. The verification system of claim 25.

27. A registration information generation method executed by a registration information generation device, which generates registration information based on registration side input information input for registration, the registration information including at least one of an element corresponding to a first set and an element corresponding to a second set different from the first set, the registration information generation method comprising: generating registration side intermediate information from the registration side input information by performing a registration side sharing process for secretly sharing elements corresponding to the first set of input information and a registration side copying process for copying elements corresponding to the second set of input information; performing a registration side replacement process for replacing the order of elements included in the input information by using a replacement key; generating the registration information by replacing the order of elements included in the registration side intermediate information by using the replacement key; How registration information is generated.

28. A registration information generation method executed by a registration information generation device, which generates registration information based on registration side input information input for registration, comprising the steps of: generating secret sharing information from the registration side input information by performing a sharing process for secretly sharing elements included in the input information; performing a registration side replacement process for replacing the order of elements included in the input information by using a replacement key; generating the registration information by replacing the order of elements included in the secret sharing information with the replacement key; How registration information is generated.

29. A matching information generation method executed by a matching information generation device, which generates matching information based on matching side input information input for matching, the matching information including at least one of elements corresponding to a first set and elements corresponding to a second set different from the first set, the matching information generation method comprising: generating verification-side intermediate information from the verification-side input information by performing a verification-side duplication process for duplication of elements corresponding to the first set of input information and a verification-side sharing process for secretly sharing elements corresponding to the second set of input information; performing a matching side permutation process for permuting the order of elements included in the input information by using a permutation key; generating the matching information by replacing the order of elements included in the matching intermediate information by using the replacement key; How matching information is generated.

30. A matching information generation method executed by a matching information generation device, which generates matching information based on matching side input information input for matching, comprising the steps of: generating duplicate information from the input information on the verification side by performing a duplication process of duplication of elements included in the input information; performing a matching side permutation process for permuting the order of elements included in the input information by using a permutation key; generating the verification information by replacing the order of elements included in the duplicate information with the replacement key; How matching information is generated.

31. A similarity calculation method for a verification system including a registration information generation device that generates registration information based on registration side input information input for registration, a verification information generation device that generates verification information based on verification side input information input for verification, and a verification device that verifies the registration side input information and the verification side input information, comprising: the registration-side input information and the matching-side input information each include at least one of an element corresponding to a first set and an element corresponding to a second set, The registration information generating device, generating registration side intermediate information from the registration side input information by performing a registration side sharing process for secretly sharing elements corresponding to the first set of input information and a registration side copying process for copying elements corresponding to the second set of input information; performing a registration side replacement process for replacing the order of elements included in the input information by using a replacement key; generating the registration information by replacing the order of elements included in the registration side intermediate information with the replacement key; The matching information generating device, generating verification-side intermediate information from the verification-side input information by performing a verification-side duplication process for duplication of elements corresponding to the first set of input information and a verification-side sharing process for secretly sharing elements corresponding to the second set of input information; performing a matching-side permutation process for permuting the order of elements included in the input information by using the permutation key; generating the verification information by replacing an order of elements included in the verification side intermediate information with the replacement key; The collation device, calculating a similarity between the registration side input information and the matching side input information based on the registration information and the matching information; Similarity calculation method.

32. A similarity calculation method for a verification system including a registration information generation device that generates registration information based on registration side input information input for registration, a verification information generation device that generates verification information based on verification side input information input for verification, and a verification device that verifies the registration side input information and the verification side input information, comprising: The registration information generating device, generating shared information from the registration side input information by performing a sharing process for secretly sharing elements included in the input information; performing a registration side replacement process for replacing the order of elements included in the input information by using a replacement key; generating the registration information by replacing an order of elements included in the shared information by using the replacement key; The matching information generating device, generating duplicate information from the input information on the verification side by performing a duplication process of duplication of elements included in the input information; performing a matching-side permutation process for permuting the order of elements included in the input information by using the permutation key; generating the verification information by replacing an order of elements included in the duplicate information with the replacement key; The collation device, calculating a similarity between the registration side input information and the matching side input information based on the registration information and the matching information; Similarity calculation method.

33. A registration information generation program that generates registration information based on registration side input information input for registration, the registration information including at least one of an element corresponding to a first set and an element corresponding to a second set different from the first set, generating registration side intermediate information from the registration side input information by performing a registration side sharing process for secretly sharing elements corresponding to the first set of input information and a registration side copying process for copying elements corresponding to the second set of input information; performing a registration side replacement process for replacing the order of elements included in the input information by using a replacement key; generating the registration information by replacing the order of elements included in the registration side intermediate information by using the replacement key; Registration information generator.

34. A registration information generation program for generating registration information based on registration side input information input for registration, generating secret sharing information from the registration side input information by performing a sharing process for secretly sharing elements included in the input information; performing a registration side replacement process for replacing the order of elements included in the input information by using a replacement key; generating the registration information by replacing the order of elements included in the secret sharing information with the replacement key; Registration information generator.

35. A matching information generating method for generating matching information based on matching side input information input for matching, the matching information including at least one of an element corresponding to a first set and an element corresponding to a second set different from the first set, the matching information comprising: generating verification-side intermediate information from the verification-side input information by performing a verification-side duplication process for duplication of elements corresponding to the first set of input information and a verification-side sharing process for secretly sharing elements corresponding to the second set of input information; performing a matching side permutation process for permuting the order of elements included in the input information using a permutation key; generating the matching information by replacing the order of elements included in the matching intermediate information by using the replacement key; Matching information generator.

36. A verification information generating program for generating verification information based on verification side input information input for verification, generating duplicate information from the input information on the verification side by performing a duplication process of duplication of elements included in the input information; performing a matching side permutation process for permuting the order of elements included in the input information using a permutation key; generating the verification information by replacing the order of elements included in the duplicate information with the replacement key; Matching information generator.

Images