Method, system and computer program product for multiple biometric template screening

• US 20040230810A1
• Filed: 05/15/2003
• Published: 11/18/2004
• Est. Priority Date: 05/15/2003
• Status: Active Grant

• Current Assignee: Activcard Ireland, Limited
• Original Assignee: Activcard Ireland, Limited
• Inventors: Hillhouse, Robert D.
• Granted Patent: US 8,171,304 B2
• Time in Patent Office: Days
• US Class Current: 713/186
• CPC Class Codes: G06V 40/1365 Matching; Classification

Abstract

A method, system and computer program product which allows identification of an enrollment biometric template having a highest probability of matching a sample biometric template from a plurality of enrolled biometric templates without compromising or significantly compromising system security. In one embodiment of the invention, first feature set information is derived from sample and enrollment biometric templates. The first feature set information generally comprises spatially dependent information associated with a fingerprint. The first feature set information is then used to determine which enrollment biometric template has the highest probability of matching the sample biometric template. Second feature set information is then derived from the biometric sample template and the determined enrollment biometric template and used to perform a one-to-one match. The second feature set information generally comprises pattern dependent information associated with a fingerprint.

Claims

1. A fingerprint biometric security method which utilizes a first feature set to screen a plurality of enrollment biometric templates for a most relevant enrollment biometric template and a second feature set for matching said most relevant enrollment biometric template comprising the steps of;

a. deriving from a biometric sample, a sample first feature set,

b. providing a plurality of enrollment first feature sets derived from each member of said plurality of enrollment biometric templates,

c. comparing in a one-to-many relationship, said sample first feature set to said plurality of enrollment first feature sets, and

d. determining said most relevant enrollment biometric template by selecting an enrollment first feature set which most favorably compares to said sample first feature set.

2. The method according to claim 1 further including the steps of;

a. deriving from said biometric sample, a sample second feature set,

b. providing an enrollment second feature set derived from said most relevant enrollment biometric template, and

c. matching said sample second feature set to said enrollment second feature set.

3. The method according to claim 1 wherein said sample first feature set is independent of said sample second feature set.

4. The method according to claim 1 wherein said sample first feature set overlaps with said sample second feature set and at least a portion of said sample second feature set is incorporated into said sample first feature set.

5. The method according to claim 1 wherein said sample first feature set is generally comprised of spatially dependent features.

6. The method according to claim 5 wherein said spatially dependent features includes minutia points or ridge spacing frequencies.

7. The method according to claim 1 wherein said sample second feature set is generally comprised of pattern dependent features.

8. The method according to claim 7 wherein said pattern dependent features include minutia types, ridge flow angles, minutia direction, ridge count or ridge patterns.

9. The method according to claim 1 wherein step 1.b further includes the step of retrieving said plurality of enrollment first feature sets from a storage location.

10. The method according to claim 1 wherein step 1.b further includes the step of deriving said plurality of enrollment first feature sets on demand.

11. The method according to claim 2 wherein step 2.b further includes the step of retrieving said enrollment second feature set from a storage location.

12. The method according to claim 2 wherein step 2.b further includes the step of deriving said enrollment second feature set on demand.

13. The method according to claim 1 wherein steps 1.a and 1.b are performed by a computer system.

14. The method according to claim 1 wherein steps 1.c and 1.d are performed by a security token in processing communications with a computer system.

15. The method according to claim 13 wherein said computer system is either a local computer system or a remote computer system.

16. The method according to claim 2 wherein said sample first feature set and said sample second feature set derived from said biometric sample are incorporated into a sample biometric template.

17. A fingerprint biometric security method which utilizes a generally spatially dependent feature set to screen a plurality of enrollment biometric templates for a most relevant enrollment biometric template and a generally pattern dependent feature set for matching said most relevant enrollment biometric template comprising the steps:

a. deriving from a biometric sample, a sample-based spatially dependent feature set,

b. deriving from each member of said plurality of enrollment biometric templates, a plurality of enrollment-based spatially dependent feature sets,

c. comparing in a one-to-many relationship, said sample-based spatially dependent feature set to said plurality of enrollment-based spatially dependent feature sets, and

d. determining said most relevant enrollment biometric template by selecting an enrollment-based spatially dependent feature set which most favorably compares to said sample-based spatially dependent feature set.

18. The method according to claim 17 further including the steps of;

a. deriving a sample-based pattern dependent feature set from said biometric sample,

b. deriving an enrollment-based pattern dependent feature set from said most relevant enrollment biometric template, and

c. matching said sample-based pattern dependent feature set to said enrollment-based pattern dependent feature set.

19. The method according to claim 17 wherein said generally spatially dependent feature set is independent of said generally pattern dependent feature set.

20. The method according to claim 17 wherein said generally spatially dependent feature set incorporates at least a portion of said generally pattern dependent feature set.

21. The method according to claim 17 further including the step of incorporating said sample-based spatially dependent feature set into a sample pre-match template.

22. The method according to claim 17 further including the step of incorporating said plurality of enrollment-based spatially dependent feature sets into a plurality of retrievable enrollment pre-match templates.

23. A fingerprint biometric security method which primarily utilizes a spatially dependent feature set to screen a plurality of enrollment biometric templates for a most relevant enrollment biometric template and a pattern dependent feature set for matching said most relevant enrollment biometric template comprising the steps:

a. deriving from a biometric sample, a sample-based feature set, wherein said sample-based feature set incorporates both spatially dependent and pattern dependent features,

b. deriving from each member of said plurality of enrollment biometric templates, a plurality of enrollment-based feature sets, wherein each set of said plurality of enrollment-based feature sets incorporates both spatially dependent and pattern dependent features,

c. comparing in a one-to-many relationship, said sample-based feature set to said plurality of enrollment-based feature sets, and

d. determining said most relevant enrollment biometric template by selecting an enrollment-based feature set which most favorably compares to said sample-based feature set.

24. The method according to claim 23 further including the steps of;

a. deriving a sample-based pattern dependent feature set from said biometric sample,

b. deriving an enrollment-based pattern dependent feature set from said most relevant enrollment biometric template, and

c. matching said sample-based pattern dependent feature set to said enrollment-based pattern dependent feature set.

25. The method according to claim 23 wherein said sample-based feature set comprises minutia points or ridge spacing frequencies and at least one pattern dependent feature selected from the group consisting of minutia types, ridge flow angles, minutia direction, ridge count and ridge patterns.

26. A fingerprint biometric security system which utilizes a first feature set to screen a plurality of enrollment biometric templates for a most relevant enrollment biometric template and a second feature set for matching said most relevant enrollment biometric template comprising;

a computer system including means for;

receiving a biometric sample,

deriving a sample first feature set from said biometric sample,

deriving a plurality of enrollment first feature sets from each member of said plurality of enrollment biometric templates,

comparing said sample first feature set to said plurality of enrollment first feature sets, and

determining said most relevant enrollment biometric template.

27. The system according to claim 26 further including means for;

deriving a sample second feature set from said biometric sample,

deriving an enrollment second feature set from said most relevant enrollment biometric template, and

matching said enrollment second feature set to said sample second feature set.

28. The system according to claim 26 wherein said first feature set is independent of said second feature set.

29. The system according to claim 26 wherein said first feature set overlaps with said second feature set.

30. The system according to claim 26 wherein said first feature set is generally comprised of spatially dependent features.

31. The system according to claim 30 wherein said spatially dependent features includes minutia points or ridge spacing frequencies.

32. The method according to claim 26 wherein said second feature set is generally comprised of pattern dependent features.

33. The system according to claim 32 wherein said pattern dependent features include minutia types, ridge flow angles, minutia direction, ridge count or ridge patterns.

34. The system according to claim 26 wherein said plurality of enrollment first feature sets are derived in advance and incorporated into enrollment pre-match templates.

35. A fingerprint biometric security system which utilizes a first feature set to screen a plurality of enrollment biometric templates for a most relevant enrollment biometric template and a second feature set for matching said most relevant enrollment biometric template comprising;

a computer system in processing communications with a security token and including means for;

receiving a biometric sample,

generating a sample biometric template which incorporates a sample first feature set and a sample second feature set, and

sending said sample biometric template to said security token.

36. The system according to claim 35 wherein said security token further includes said plurality of enrollment biometric templates operatively stored therein and further includes means for;

receiving said sample biometric template,

deriving an enrollment first feature set from each member of said plurality of enrollment biometric templates,

comparing said sample first feature set to each derived enrollment first feature set, and

determining said most relevant enrollment biometric template.

37. The system according to claim 36 wherein said security token further includes means for;

deriving a most relevant second feature set from said most relevant enrollment biometric template, and

matching said sample second feature set to said most relevant second feature set.

38. The system according to claim 36 wherein said security token having operatively stored therein;

said plurality of enrollment biometric template, and

a plurality of pre-match enrollment templates derived from and relationally associated with said plurality of enrollment biometric templates; and

said security token further including means for;

receiving said sample biometric template,

comparing said sample first feature set to said plurality of pre-match enrollment templates, and

determining said most relevant enrollment biometric template.

39. The system according to claim 38 wherein said security token further includes means for;

deriving a sample second feature set from said sample biometric template,

deriving a most relevant second feature set from said most relevant enrollment biometric template, and

matching said sample second feature set to said most relevant second feature set.

40. A computer program product embodied in a tangible form readable by a processor having executable institutions stored thereon for causing a computer system to screen a plurality of enrollment biometric templates for a most relevant enrollment biometric template using a first feature set and to match said most relevant enrollment biometric template using a second feature set, comprising the executable instructions to;

a. derive from a biometric sample, a sample first feature set,

b. derive a plurality of enrollment first feature sets from each member of said plurality of enrollment biometric templates,

c. compare in a one-to-many relationship, said sample first feature set to said plurality of enrollment first feature sets, and

d. determine said most relevant enrollment biometric template by selection of an enrollment first feature set which most favorably compares to said sample first feature set.

41. The computer program product according to claim 40 further including the executable instructions to;

a. derive from said biometric sample, a second feature set,

b. derive an enrollment second feature set from said most relevant enrollment biometric template, and

c. match said second feature set to said enrollment second feature set.

42. A computer program product embodied in a tangible form readable by a processor having executable instructions stored thereon for causing a computer system to screen a plurality of enrollment biometric templates for a most relevant enrollment biometric template using a first feature set and to match said most relevant enrollment biometric template using a second feature set, comprising the executable instructions to;

a. derive from a biometric sample, a sample-based feature set, wherein said sample-based feature set incorporates both spatially dependent and pattern dependent features,

b. derive from each member of said plurality of potentially relevant enrollment biometric templates, a plurality of enrollment-based feature sets, wherein each set of said plurality of enrollment-based feature sets incorporates both spatially dependent and pattern dependent features,

c. compare in a one-to-many relationship, said sample-based set to said plurality of enrollment-based feature sets, and

d. determine said most relevant enrollment biometric template by selection of an enrollment-based feature set which most favorably compares to said sample-based feature set.

43. A method according to claim 42, wherein the a second criteria at least partially independent of the first criteria such that the cumulative false acceptance rate resulting from each of the registration process and the search is different than the false acceptance rate of either one of the registration process and the search in isolation.

44. A method according to claim 42, wherein the registration process includes the steps of:

determining from the first biometric information sample at least a feature within the first biometric information sample; and

,

comparing data relating to the at least a feature with template data within the intermediary biometric template to result in a registration result.

42-1. A method comprising the steps of:

a) providing a plurality of first template data relating to each of a plurality of different biometric information samples and associated therewith;

b) providing a plurality of second biometric templates, each second biometric template associated with at least one of the biometric information samples relating to the first template data;

c) providing a first biometric information sample;

d) deriving a first criteria from the first biometric information sample;

e) using the first criteria, searching the plurality of first template data to determine at least an intermediary biometric template from the plurality of second biometric template with which a likelihood of registration against the biometric information sample is above a predetermined threshold;

f) for each of the at least an intermediary biometric template performing a registration against the first biometric information sample according to a registration process, the registration process dependent upon a second criteria at least partially independent of the first criteria; and

,

g) in dependence upon a result of the registration, determining a resultant biometric template from the at least an intermediary biometric template with which the first biometric information sample registers beyond a predetermined threshold of likelihood.