Method and system for adaptively varying templates to accommodate changes in biometric information

• US 7,103,200 B2
• Filed: 03/05/2001
• Issued: 09/05/2006
• Est. Priority Date: 03/05/2001
• Status: Expired due to Term

• Current Assignee: Activcard Ireland, Limited (Assa Abloy AB)
• Original Assignee: Activcard Ireland, Limited (Assa Abloy AB)
• Inventors: Boyer, John, Hillhouse, Robert
• Primary Examiner(s): Chawan, Sheela
• Application Number:
  US09/797,975
• Publication Number:
  US 20020154793A1
• Time in Patent Office: 2,010 Days
• Field of Search: 382/115, 382/124, 382/278, 382/128, 382/209, 382/312, 235/380, 235/376, 380/255, 380/258, 713/186, 713/172, 713/182, 713/185
• US Class Current: 382/115
• CPC Class Codes: G06F 18/28 Determining representative reference patterns, e.g. by averaging or distorting; Generating dictionaries
• CPC Class Codes: G06V 40/1365 Matching; Classification

Abstract

A method of authenticating a user in dependence upon biometric input information is disclosed. According to the method when a user is identified, their biometric information or data derived therefrom is automatically stored. The data is then used in subsequent user identification attempts as a template. The original templates are not replaced but those templates that are automatically stored are replaced at intervals.

Claims

1. A method of identifying a user comprising the steps of:

a) receiving a contact image biometric information sample from a biometric source of the user for use in identifying the user;

b) determining electronic data representing said contact image biometric information sample of the user;

c) analyzing the electronic data to compare the electronic data against biometric templates relating to identifications of users;

d) when a first threshold corresponding to a substantial match of a template is satisfied, identifying the individual; and

,

e) when a difference between the electronic data and the matched template is above a predetermined second threshold, storing a new biometric template as an adaptive enrollment for the user based on the electronic data relating to biometric information samples for said biometric source as provided by the user for use in identifying the user by (a) through (c).

2. A method according to claim 1 wherein every predetermined number of occurrences of a substantial match:

the electronic data is stored within a history file;

the electronic data is analyzed to determine its suitability for use as the adaptive enrollment; and

,

when the electronic data is determined to be suitable, storing the electronic data as the adaptive enrollment.

3. A method according to claim 2 wherein the history file comprises up to N samples of electronic data and wherein an oldest sample of electronic data is deleted when a new sample of electronic data is added to a history file containing N samples.

4. A method according to claim 3 wherein a score is stored in association with each sample electronic data within the history file indicating a suitability of the sample for use as the adaptive enrollment.

5. A method according to claim 4 wherein the score is determined by comparing the sample electronic data against each other sample electronic data within the history file and associated with a same user identification.

6. A method according to claim 5 wherein the score is determined by determining a composite comparison metric, M_i for each sample electronic data, i,

$m_i=\frac{\sum _{}^{}_{}}{}$

j = 1 n j ≠

1 ⁢

⁢

c ⁡

( H i , H j ) n - 1 , i ∈

{ 1 , …

⁢

, n }

where n is the number of historical enrollments in the set and c(H_i,H_j) is the comparison metric between biometric template i and biometric template j.

7. A method according to claim 5 wherein upon deleting a sample electronic data from the history file, its effect on the scores in the history file is approximately reversed.

8. A method according to claim 7 wherein the effect is approximately reversed by recalculating m, after the deletion of sample electronic element j according to

$m_i=m_i-\frac{c}{}$

( H i , H j ) n - 1 , i ∈

{ 1 , …

⁢

, n } , i ≠

j

where n is the number of historical enrollments in the set including historical enrollment j.

9. A method according to claim 8 wherein the effect of an added sample electronic element j is determined by calculating m_i according to

$m_i=m_i+\frac{c}{}$

( H i , H j ) n - 1 , i ∈

{ 1 , …

⁢

, n } , i ≠

j

where n is the number of historical enrollments in the set including historical enrollment j.

10. A method according to claim 2 wherein the effect of an added sample electronic element j is determined by calculating m_i according to

$m_i=m_i+\frac{c}{}$

( H i , H j ) n - 1 , i ∈

{ 1 , …

⁢

, n } , i ≠

j

where n is the number of historical enrollments in the set including historical enrollment i.

11. A method according to claim 2 wherein electronic data samples suitable for use as the adaptive enrollment are compared against the stored data, and when a difference greater than a predetermined difference exists, a notification of such is generated.

12. A method according to claim 2 wherein electronic data samples suitable for use in determining the other stored data are compared against the stored data, and when a difference greater than a predetermined difference exists, the user is prompted to re-enroll in order to provide new biometric information for use in determining new stored data.

13. A method of identifying a user comprising the steps of:

a) receiving a contact image biometric information sample from a biometric source of the user for use in identifying the user;

b) determining electronic data representing said contact image biometric information sample of the user;

c) analyzing the electronic data to compare the electronic data against stored biometric template data relating to identifications of users;

d) when a first threshold corresponding to a substantial match of a template is satisfied;

identifying the individual;

storing the electronic data in association with the user identification as further stored biometric template data;

determining from the further stored biometric template data comparison data for use in identifying the user; and

,

e) when a difference between the electronic data and the stored biometric template data is above a predetermined second threshold, comparing the electronic data against other stored biometric template data relating to contact image biometric information samples for said biometric source as provided for use in identifying the user according to (a) to(c).

14. A method according to claim 13 wherein the further stored biometric template data relates to different features of the individual, the features for use in identifying the individual.

15. A method according to claim 13 wherein the step of storing the electronic data in association with the user identification as further stored biometric template data comprises:

storing an electronic representation of the biometric information sample associated with the identification.

16. A method according to claim 15 wherein determining from the further stored biometric template data comparison data for use in identifying the user is performed by:

comparing each stored electronic representation associated with the identification against a template derived from each other electronic representation associated with the identification to select at least an electronic representation that best represents the stored electronic representations associated with the identification; and

,

determining at least a template from the at least a selected electronic representation.

17. A method according to claim 15 wherein determining from the further stored biometric template data comparison data for use in identifying the user is performed by:

storing with each electronic representation associated with the identification a score indicative of a quality of the stored electronic representations associated with the identification for use in generating a template;

comparing the oldest stored electronic representation associated with the identification against a template derived from each other electronic representation associated with the identification to determine the correction for the score associated with said template;

correcting each score in accordance with the determined corrections;

comparing the newest stored electronic representation associated with the identification against a template derived from each other electronic representation associated with the identification to determine a second correction for each score;

correcting each score in accordance with the determined second correction; and

,

selecting an electronic representation associated with a best score for use in determining a template for use as the comparison data.

18. A method of identifying a user comprising:

receiving a contact image biometric information sample from a biometric source of the user for use in identifying the user;

determining electronic data representing said contact image biometric information sample of the user;

analyzing the electronic data to extract features for comparison with stored electronic data;

comparing the extracted features against stored electronic data, the stored electronic data relating to identifications of users;

when a substantial match is found, performing;

identifying the individual;

determining a number of times that the user has been identified and when the number is one of at least a predetermined number;

storing the electronic data in association with the user identification as further stored data;

determining from the further stored data comparison data for use in identifying the user; and

,

when a difference between the electronic data and the stored data is above a predetermined threshold, comparing the electronic data against other stored data, the other stored data relating to additional biometric information samples for said biometric source as provided for use in identifying the user.

19. A method of identifying a user comprising:

receiving a biometric information sample from a biometric source using a contact imaging device that images the biometric source to provide said biometric information sample;

forming from said biometric information sample at least a first template including biometric features for use in identifying an individual;

forming a history file including a plurality of biometric information samples from said biometric source each provided by the same individual at different times for use in identifying the individual; and

forming at least one adaptive enrollment template from the biometric information samples from said biometric source in the history file,

wherein an individual is identified by registering provided biometric information sample against the at least a first template and, when registration is not successful, registering the provided biometric information sample against the at least one adaptive enrollment template.

20. A method according to claim 19 wherein the at least one adaptive enrollment template is a template formed from one or more biometric information samples within the history file, the one or more biometric information samples selected through a comparison with other biometric information samples within the history file.

21. A method according to claim 20 wherein each biometric information sample stored within the history file is compared against each other biometric information sample within the history file to determine those that are most suitable for use in determining adaptive enrollment templates.

22. A method according to claim 21 wherein the history file comprises up to N biometric information samples and wherein an oldest biometric information sample is deleted when a new biometric information sample is added to a history file containing N samples.

23. A method according to claim 22 wherein a score is stored in association with each biometric information sample within the history file indicating a suitability of that sample for use in determining an adaptive enrollment template.

24. A method according to claim 23 wherein the score is determined by comparing the sample electronic data against each other sample electronic data within the history file and associated with a same user identification.

25. A method according to claim 24 wherein the score is determined by determining a composite comparison metric, M_i for each sample electronic data, i,

$m_i=\frac{\sum _{}^{}_{}}{}$

j = 1 n j ≠

1 ⁢

⁢

c ⁡

( H i , H j ) n - 1 , i ∈

{ 1 , …

⁢

, n }

where n is the number of historical enrollments in the set and c(H_i,H_j) is the comparison metric between biometric template i and biometric template j.

26. A method according to claim 24 wherein upon deleting a sample electronic data from the history file, its effect on the scores in the history file is approximately reversed.

27. A method according to claim 26 wherein the effect is approximately reversed by recalculating m_i after the deletion of sample electronic element j according to

$m_i=m_i-\frac{c}{}$

( H i , H j ) n - 1 , i ∈

{ 1 , …

⁢

, n } , i ≠

j

where n is the number of historical enrollments in the set including historical enrollment j.

28. A method according to claim 27 wherein the effect of an added sample electronic element j is determined by calculating m_i according to

$m_i=m_i+\frac{c}{}$

( H i , H j ) n - 1 , i ∈

{ 1 , …

⁢

, n } , i ≠

j

wherein n is the number of historical enrollments in the set including historical enrollment j.

29. A method according to claim 28 wherein the effect of an added sample electronic element j is determined by calculating m_i according to

$m_i=m_i+\frac{c}{}$

( H i , H j ) n - 1 , i ∈

{ 1 , …

⁢

, n } , i ≠

j

where n is the number of historical enrollments in the set including historical enrollment j.

30. A method according to claim 19 wherein adaptive enrollment templates are compared against the at least a first template and, when a difference greater than a predetermined difference exists, a notification of such is generated.

31. A method according to claim 19 wherein adaptive enrollment templates are compared against the at least a first template and, when a difference greater than a predetermined difference exists, the user is prompted to re-enroll in order to provide new biometric information for use in determining a new at least a first template.