Method and device for authenticating identify by means of fusion of multiple biological characteristics

• US 10,346,602 B2
• Filed: 03/29/2016
• Issued: 07/09/2019
• Est. Priority Date: 10/20/2015
• Status: Active Grant

• Current Assignee: GRG Banking Equipment Company Limited (City of Guangzhou)
• Original Assignee: GRG Banking Equipment Company Limited (City of Guangzhou)
• Inventors: Xiao, Zheng, Zheng, Jiachun, Han, Xiaoping, He, Jinjun
• Primary Examiner(s): Gee, Jason K
• Assistant Examiner(s): Almamun, Abdullah
• Application Number:
  US15/765,253
• Publication Number:
  US 20180285542A1
• Time in Patent Office: 1,197 Days
• CPC Class Codes: G06F 18/253 of extracted features
• CPC Class Codes: G06F 21/32 using biometric data, e.g. fingerprints, iris scans or voiceprints
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks
• CPC Class Codes: G06V 10/806 of extracted features
• CPC Class Codes: G06V 40/70 Multimodal biometrics, e.g. combining information from different biometric modalities
• CPC Class Codes: H04L 63/0861 using biometrical features, e.g. fingerprint, retina-scan cryptographic mechanisms or cryptographic arrangements for entity authentication using biological data H04L9/3231

Abstract

Provided are a method and device for authenticating an identity based on fusion of multiple biological characteristics. The method includes: collecting at least two types of biological characteristic identity information of a to-be-identified user; performing characteristic extraction on each type of the collected biological characteristic identify information, to obtain characteristic information corresponding to the type; establishing characteristic matrixes based on the characteristic information; performing normalization processing on each of the characteristic matrixes; performing dynamic weighting fusion on all of the normalized characteristic matrixes, to obtain a fused characteristic matrix; matching the fused characteristic matrix with a preset standard matrix, to obtain a matching score; and obtaining an identity identification result of the to-be-identified user based on a Bayesian decision model and the matching score.

Claims

1. A method for authenticating an identity based on fusion of a plurality of biological characteristics, comprising:

collecting at least two types of biological characteristic identity information of a to-be-identified user;

performing characteristic extraction on each type of the at least two types of biological characteristic identify information, to obtain sets of characteristic information, wherein there is a one-to-one correspondence between the sets of characteristic information and the at least two types of biological characteristic identify information;

establishing characteristic matrixes based on the sets of characteristic information, wherein there is another one-to-one correspondence between the sets of characteristic information and the characteristic matrixes;

performing normalization processing on each of the characteristic matrixes;

performing dynamic weighting fusion on all of the normalized characteristic matrixes, to obtain a fused characteristic matrix;

matching the fused characteristic matrix with a preset standard matrix, to obtain a matching score; and

obtaining an identity identification result of the to-be-identified user based on a Bayesian decision model and the matching score;

wherein the performing dynamic weighting fusion on all of the normalized characteristic matrixes to obtain a fused characteristic matrix comprises;

allocating an initial weighting value to each of the normalized characteristic matrixes according to a preset weighting value allocation criteria, wherein a sum of all of the initial weighting values is equal to 1;

calculating an integrity rate for each of the sets of characteristic information corresponding to the characteristic matrixes, wherein the integrity rate is a ratio of a quantity of elements in the set of characteristic information to a quantity of elements in the corresponding characteristic matrix;

adjusting the initial weighting value of each of the normalized characteristic matrixes based on the integrity rates while ensuring a sum of all of the adjusted weighting values is equal to 1; and

performing weighting fusion on all of the normalized characteristic matrixes based on the adjusted weighting values, to obtain the fused characteristic matrix.

2. The method according to claim 1, wherein the performing normalization processing on each of the characteristic matrixes comprises:

performing normalization processing on each of the characteristic matrixes using a MAX-MIN method, wherein each element in each of the normalized characteristic matrixes ranges from 0 to 1, inclusively.

3. The method according to claim 2, wherein the standard matrix is reserved before an account is logged into for the first time using the method, and the standard matrix is reserved by:

collecting at least two types of biological characteristic identity information of an account holder;

performing characteristic extraction on each type of the at least two types of collected biological characteristic identity information of the account holder, to obtain sets of characteristic information of the account holder, wherein there is a one-to-one correspondence between the sets of characteristic information of the account holder and the at least two types of biological characteristic identify information of the account holder;

establishing characteristic matrixes of the account holder based on the sets of characteristic information of the account holder, wherein there is another one-to-one correspondence between the sets of characteristic information of the account holder and the characteristic matrixes of the account holder;

performing the normalization processing on each of the characteristic matrixes of the account holder;

performing the dynamic weighting fusion on all of the normalized characteristic matrixes of the account holder, to obtain a fused characteristic matrix corresponding to the account holder; and

storing the fused characteristic matrix corresponding to the account holder in a database as the standard matrix.

4. The method according to claim 1, wherein the adjusting the weighting value of each of the characteristic matrixes based on the integrity rates while ensuring a sum of all of the weighting values is equal to 1 comprises:

in a case that the integrity rate corresponding to one of the characteristic matrixes is less than 1, proportionally decreasing the initial weighting value of the one of the characteristic matrixes based on a missing rate corresponding to the one of the characteristic matrixes, wherein a sum of the integrity rate and the missing rate is equal to 1;

in a case that the integrity rate corresponding to one of the characteristic matrixes is equal to 1, increasing the initial weighting value of the one of the characteristic matrixes corresponding to the integrity rate equal to 1 based on a proportion of the initial weighting value of the one of the characteristic matrixes, when the initial weighting value of another one of the characteristic matrixes corresponding to the integrity rate less than 1 is decreased proportionally, while ensuring the sum of all of the weighting values is equal to 1;

in a case that each of the integrity rates corresponding to the characteristic matrixes is less than 1, regarding one of the characteristic matrixes corresponding to the integrity rate greater than a preset threshold as a characteristic matrix corresponding to the integrity rate equal to 1; and

in a case that each of the integrity rates corresponding to the characteristic matrixes is less than or equal to the preset threshold, returning to the step of collecting at least two types of biological characteristic identity information of the to-be-identified user, for re-collection.

5. The method according to claim 4, wherein the standard matrix is reserved before an account is logged into for the first time using the method, and the standard matrix is reserved by:

collecting at least two types of biological characteristic identity information of an account holder;

performing characteristic extraction on each type of the at least two types of collected biological characteristic identity information of the account holder, to obtain sets of characteristic information of the account holder, wherein there is a one-to-one correspondence between the sets of characteristic information of the account holder and the at least two types of biological characteristic identify information of the account holder;

establishing characteristic matrixes of the account holder based on the sets of characteristic information of the account holder, wherein there is another one-to-one correspondence between the sets of characteristic information of the account holder and the characteristic matrixes of the account holder;

performing the normalization processing on each of the characteristic matrixes of the account holder;

performing the dynamic weighting fusion on all of the normalized characteristic matrixes of the account holder, to obtain a fused characteristic matrix corresponding to the account holder; and

storing the fused characteristic matrix corresponding to the account holder in a database as the standard matrix.

6. The method according to claim 1, wherein the to-be-identified user is allowed to handle business on a currently-logged account in a case that the identity identification result indicates that the to-be-identified user passes identity identification;

and the method returns to the step of collecting at least two types of biological characteristic identity information of the to-be-identified user and an alarm is raised to the logged account, in a case that the identity identification result indicates that the to-be-identified user fails to pass the identity identification.

7. The method according to claim 6, wherein the standard matrix is reserved before an account is logged into for the first time using the method, and the standard matrix is reserved by:

collecting at least two types of biological characteristic identity information of an account holder;

performing characteristic extraction on each type of the at least two types of collected biological characteristic identity information of the account holder, to obtain sets of characteristic information of the account holder, wherein there is a one-to-one correspondence between the sets of characteristic information of the account holder and the at least two types of biological characteristic identify information of the account holder;

establishing characteristic matrixes of the account holder based on the sets of characteristic information of the account holder, wherein there is another one-to-one correspondence between the sets of characteristic information of the account holder and the characteristic matrixes of the account holder;

performing the normalization processing on each of the characteristic matrixes of the account holder;

performing the dynamic weighting fusion on all of the normalized characteristic matrixes of the account holder, to obtain a fused characteristic matrix corresponding to the account holder; and

storing the fused characteristic matrix corresponding to the account holder in a database as the standard matrix.

8. The method according to claim 1, wherein the standard matrix is reserved before an account is logged into for the first time using the method, and the standard matrix is reserved by:

collecting at least two types of biological characteristic identity information of an account holder;

performing characteristic extraction on each type of the at least two types of collected biological characteristic identity information of the account holder, to obtain sets of characteristic information of the account holder, wherein there is a one-to-one correspondence between the sets of characteristic information of the account holder and the at least two types of biological characteristic identify information of the account holder;

establishing characteristic matrixes of the account holder based on the sets of characteristic information of the account holder, wherein there is another one-to-one correspondence between the sets of characteristic information of the account holder and the characteristic matrixes of the account holder;

performing the normalization processing on each of the characteristic matrixes of the account holder;

performing the dynamic weighting fusion on all of the normalized characteristic matrixes of the account holder, to obtain a fused characteristic matrix corresponding to the account holder; and

storing the fused characteristic matrix corresponding to the account holder in a database as the standard matrix.

9. A device for authenticating an identity based on fusion of a plurality of biological characteristics, comprising:

an identity information collecting module configured to collect at least two types of biological characteristic identity information of a to-be-identified user;

a characteristic extracting module configured to perform characteristic extraction on each type of the at least two types of biological characteristic identify information, to obtain sets of characteristic information, wherein there is a one-to-one correspondence between the sets of characteristic information and the at least two types of biological characteristic identify information;

a matrix establishing module configured to establish characteristic matrixes based on the sets of characteristic information, wherein there is another one-to-one correspondence between the sets of characteristic information and the characteristic matrixes;

a normalization processing module configured to perform normalization processing on each of the characteristic matrixes;

a fusing matrix module configured to perform dynamic weighting fusion on all of the normalized characteristic matrixes, to obtain a fused characteristic matrix;

a matching module configured to match the fused characteristic matrix with a preset standard matrix, to obtain a matching score; and

an identifying module configured to obtain an identity identification result of the to-be-identified user based on a Bayesian decision model and the matching score;

wherein the fusing matrix module comprises;

an initial weighting value allocating unit configured to allocate an initial weighting value to each of the normalized characteristic matrixes according to a preset weighting value allocation criteria, wherein a sum of all of the initial weighting values is equal to 1;

an integrity rate calculating unit configured to calculate an integrity rate for each of the characteristic information corresponding to the characteristic matrixes, wherein the integrity rate is a ratio of a quantity of elements in the set of characteristic information to a quantity of elements in the corresponding characteristic matrix;

a weighting value adjusting unit configured to adjust the weighting value of each of the normalized characteristic matrixes based on the integrity rates while ensuring a sum of all of the adjusted weighting values is equal to 1; and

a weighting fusion unit configured to perform weighting fusion on all of the normalized characteristic matrixes based on the adjusted weighting values, to obtain the fused characteristic matrix.

10. The device according to claim 9, further comprising:

a business allowing module configured to allow the to-be-identified user to handle business on a currently-logged account in a case that the identity identification result obtained by the identifying module indicates that the to-be-identified user passes identity identification; and

an alarm module configured to return to the step of collecting at least two types of biological characteristic identity information of the to-be-identified user and raise an alarm to the logged account, in a case that the identity identification result obtained by the identifying module indicates that the to-be-identified user fails to pass the identity identification.

11. The device according to claim 10, further comprising a reserving module configured to reserve the standard matrix before the device is used for the first time, wherein the reserving module comprises:

an information collecting unit configured to collect at least two types of biological characteristic identity information of an account holder;

a characteristic extracting unit configured to perform characteristic extraction on each type of the at least two types of collected biological characteristic identity information of the account holder, to obtain sets of characteristic information of the account holder, wherein there is a one-to-one correspondence between the sets of characteristic information of the account holder and the at least two types of biological characteristic identify information of the account holder;

a characteristic matrix establishing unit configured to establish characteristic matrixes of the account holder based on the sets of characteristic information of the account holder, wherein there is another one-to-one correspondence between the sets of characteristic information of the account holder and the characteristic matrixes of the account holder;

a normalizing unit configured to perform the normalization processing on each of the characteristic matrixes of the account holder;

a matrix fusing unit configured to perform the dynamic weighting fusion on all of the normalized characteristic matrixes of the account holder, to obtain a fused characteristic matrix corresponding to the account holder; and

a storage unit configured to store the fused characteristic matrix corresponding to the account holder in a database as the standard matrix.

12. The device according to claim 9, further comprising a reserving module configured to reserve the standard matrix before the device is used for the first time, wherein the reserving module comprises:

an information collecting unit configured to collect at least two types of biological characteristic identity information of an account holder;

a characteristic extracting unit configured to perform characteristic extraction on each type of the at least two types of collected biological characteristic identity information of the account holder, to obtain sets of characteristic information of the account holder, wherein there is a one-to-one correspondence between the sets of characteristic information of the account holder and the at least two types of biological characteristic identify information of the account holder;

a characteristic matrix establishing unit configured to establish characteristic matrixes of the account holder based on the sets of characteristic information of the account holder, wherein there is another one-to-one correspondence between the sets of characteristic information of the account holder and the characteristic matrixes of the account holder;

a normalizing unit configured to perform the normalization processing on each of the characteristic matrixes of the account holder;

a matrix fusing unit configured to perform the dynamic weighting fusion on all of the normalized characteristic matrixes of the account holder, to obtain a fused characteristic matrix corresponding to the account holder; and

a storage unit configured to store the fused characteristic matrix corresponding to the account holder in a database as the standard matrix.