Spot-based finger biometric processing method and associated sensor

• US 7,787,667 B2
• Filed: 10/01/2004
• Issued: 08/31/2010
• Est. Priority Date: 10/01/2003
• Status: Expired due to Fees

• Current Assignee: Apple Inc.
• Original Assignee: AuthenTec Incorporated (Apple Inc.)
• Inventors: Boshra, Michael
• Primary Examiner(s): Bella; Matthew C
• Assistant Examiner(s): LIEW, ALEX KOK SOON
• Application Number:
  US10/957,331
• Publication Number:
  US 20050100199A1
• Time in Patent Office: 2,160 Days
• Field of Search: 382/124, 382/125
• US Class Current: 382/124
• CPC Class Codes: G06F 18/28 Determining representative reference patterns, e.g. by averaging or distorting; Generating dictionaries
• CPC Class Codes: G06V 10/772 Determining representative reference patterns, e.g. averaging or distorting patterns; Generating dictionaries
• CPC Class Codes: G06V 40/1376 Matching features related to ridge properties or fingerprint texture

Abstract

A method for finger biometric processing may include selecting at least one enrollment spot from finger biometric enrollment data comprising a plurality of pixels, and selecting at least one verification spot from finger biometric verification data comprising a plurality of pixels. One or more spot properties is determined for one (or both) of the at least one enrollment spot and the at least one verification spot. The method may further include comparing the at least one enrollment spot with the at least one verification spot based upon a function of corresponding pixel values of the at least one enrollment spot and the at least one verification spot, and also based upon the at least one spot property.

Claims

1. A method for finger biometric processing comprising:

sensing finger biometric verification data comprising a plurality of pixels using a finger biometric sensor;

iteratively selecting, using a processor, a sequence of enrollment spots from finger biometric enrollment data comprising a plurality of pixels representing finger ridge data;

selecting, using the processor, at least one verification spot from the finger biometric verification data comprising a plurality of pixels by

generating information content data from the finger biometric enrollment data by generating curvature data based upon the finger ridge data,

determining at least one peak in the curvature data, and

selecting the at least one enrollment spot from the finger biometric enrollment data based upon the at least one peak in the curvature data;

determining, using the processor, at least one spot property for at least one of the sequence enrollment spots and the at least one verification spot;

comparing, using the processor, at least one of the sequence of enrollment spots with the at least one verification spot based upon a function of corresponding pixel values of at least one of the sequence of enrollment spots and the at least one verification spot, and also based upon the at least one spot property;

defining, using the processor, forbidden regions surrounding the sequence of enrollment spots based upon the curvature data; and

selecting, using the processor, subsequent enrollment spots based upon the forbidden regions surrounding prior enrollment spots.

2. A method according to claim 1 wherein each forbidden region is defined by a binary function.

3. A method according to claim 1 wherein each forbidden region is defined by a penalty function.

4. A method according to claim 1 wherein the at least one spot property comprises information content data.

5. A method according to claim 1 wherein determining the at least one spot property comprises determining an uncertainty associated with at least one of the at least one enrollment spot and the at least one verification spot.

6. A method according to claim 1 further comprising generating probability distribution functions for at least one of true matches and false matches during an enrollment learning phase;

and wherein the comparison is based upon the probability distribution functions.

7. A method according to claim 6 further comprising generating a family of scoring functions to be used in the comparing based upon the probability distribution functions for at least one of true matches and false matches.

8. A method according to claim 7 wherein the family of scoring functions are based upon a conditional false acceptance rate.

9. A method according to claim 6 wherein the family of scoring functions are based upon a conditional likelihood ratio.

10. A method according to claim 1 wherein selecting the at least one enrollment spot comprises selecting a plurality of enrollment spots;

and wherein selecting the at least one verification spot comprises selecting a plurality of verification spots.

11. A method according to claim 1 wherein the finger biometric enrollment data comprises at least one of finger biometric enrollment pattern data and finger biometric enrollment minutia data.

12. A method according to claim 1 further comprising generating the finger biometric enrollment data using at least one of a finger biometric touch integrated circuit sensor and a finger biometric slide integrated circuit sensor.

13. A method for finger biometric processing comprising:

sensing finger ridge data using a finger biometric sensor;

generating, using a processor, probability distribution functions for at least one of true matches and false matches during an enrollment learning phase;

generating, using the processor, finger ridge curvature data from the finger ridge data;

determining, using the processor, at least one peak in the finger ridge curvature data;

selecting, using the processor, at least one enrollment spot from the finger ridge data based upon the at least one peak in the finger ridge curvature data;

generating, using the processor, at least one verification spot from finger biometric verification data; and

comparing, using the processor, the at least one enrollment spot with the at least one verification spot to determine a match therewith, the comparison being based upon the probability distribution functions.

14. A method according to claim 13 wherein selecting comprises iteratively selecting a sequence of enrollment spots;

wherein forbidden regions are defined surrounding the finger biometric spots based upon the finger ridge curvature data; and

wherein subsequent enrollment spots are selected based upon the forbidden regions surrounding prior enrollment spots.

15. A method according to claim 14 wherein each forbidden region is defined by at least one of a binary function and a penalty function.

16. A method according to claim 13 further comprising generating an uncertainty for the at least one enrollment spot;

and wherein comparing comprises comparing based upon the uncertainty.

17. A method according to claim 13 further comprising determining information content data for the at least one enrollment spot;

and wherein comparing comprises comparing based upon the information content data.

18. A method according to claim 13 further comprising generating a family of scoring functions to be used in the comparing based upon the probability distribution functions for at least one of true matches and false matches.

19. A method for finger biometric processing comprising:

sensing finger biometric verification data using a finger biometric sensor;

selecting, using a processor, at least one enrollment spot from the finger biometric enrollment data, each enrollment spot having a plurality of spot properties;

generating, using the processor, scoring information for the spot properties of the at least one enrollment spot by

generating probability distribution functions for at least one of true matches and false matches during an enrollment learning phase, and

generating a family of scoring functions based upon the probability distribution functions for at least one of true matches and false matches;

generating, using the processor, at least one verification spot from the finger biometric verification data; and

comparing, using the processor, the at least one enrollment spot with the at least one verification spot based upon the family of scoring functions to determine a match therewith.

20. A method according to claim 19 wherein the spot properties are independent.

21. A method according to claim 19 wherein the family of scoring functions are based upon a conditional false acceptance rate.

22. A method according to claim 19 wherein the family of scoring functions are based upon a conditional likelihood ratio.

23. A method according to claim 19 further comprising generating information content data from the finger biometric data;

and wherein selecting comprises selecting the at least one enrollment spot based upon the information content data.

24. A finger biometric sensor comprising:

a finger biometric sensing area; and

a processor connected to said finger biometric sensing area for

generating finger biometric enrollment data comprising finger ridge data and generating probability distribution functions for at least one of true matches and false matches during an enrollment learning phase,

generating information content data from the finger biometric enrollment data as finger ridge curvature data based upon the finger ridge data,

selecting at least one enrollment spot from the finger biometric enrollment data based upon the information content data,

generating an uncertainty for the at least one enrollment spot,

generating at least one verification spot from finger biometric verification data, and

comparing the at least one enrollment spot with the at least one verification spot based upon the uncertainty and the probability distribution functions to determine a match therewith; and

a memory connected to said processor for storing the at least one enrollment spot and uncertainty associated therewith.

25. A finger biometric sensor according to claim 24 wherein said processor further determines at least one peak in the finger ridge curvature data;

and wherein said processor selects the at least one enrollment spot based upon the at least one peak in the finger ridge curvature data.

26. A finger biometric sensor according to claim 24 wherein said processor iteratively selects a sequence of enrollment spots, defines forbidden regions surrounding the finger biometric spots based upon the finger ridge curvature data and with subsequent enrollment spots being selected based upon the forbidden regions surrounding prior enrollment spots.

27. A finger biometric sensor according to claim 26 wherein each forbidden region is defined by at least one of a binary function, and a penalty function.

28. A finger biometric sensor according to claim 24 wherein said processor generates the uncertainty based upon a plurality of enrollment spots.

29. A finger biometric sensor according to claim 24 wherein said processor further generates information content data for the at least one enrollment spot, and compares the at least one enrollment spot with the at least one verify spot also based upon the information content data.

30. A finger biometric sensor according to claim 24 wherein said processor further generates a family of scoring functions to be used in the comparing based upon the probability distribution functions for at least one of true matches and false matches.

31. A finger biometric sensor according to claim 24 wherein said finger biometric sensing area comprises one of an integrated circuit touch sensor area and an integrated circuit slide sensor area.

32. A finger biometric sensor comprising:

a finger biometric sensing area; and

a processor connected to said finger biometric sensing area for

generating probability distribution functions for at least one of true matches and false matches during an enrollment learning phase,

generating finger ridge data,

generating finger ridge curvature data from finger ridge data,

determining at least one peak in the finger ridge curvature data,

selecting at least one enrollment spot from the finger ridge data based upon the at least one peak in the finger ridge curvature data,

generating at least one verification spot from finger biometric verification data, and

comparing the at least one enrollment spot with the at least one verification spot to determine a match therewith, the comparison being based on the probability distribution functions.

33. A finger biometric sensor according to claim 32 wherein said processor iteratively selects a sequence of enrollment spots, defines forbidden regions surrounding the finger biometric spots based upon the finger ridge curvature data and with subsequent enrollment spots being selected based upon the forbidden regions surrounding prior enrollment spots.

34. A finger biometric sensor according to claim 33 wherein each forbidden region is defined by at least one of a binary function, and a penalty function.

35. A finger biometric sensor according to claim 33 wherein said processor further generates an uncertainty for the at least one enrollment spot;

and wherein said processor compares the at least one enrollment spot with the at least one verification spot based upon the uncertainty.

36. A finger biometric sensor according to claim 32 wherein said processor further generates a family of scoring functions to be used in the comparing based upon the probability distribution functions for at least one of true matches and false matches.

37. A finger biometric sensor comprising:

a finger biometric sensing area; and

a processor connected to said finger biometric sensing area for

generating finger biometric enrollment data,

selecting at least one enrollment spot from finger biometric enrollment data, each enrollment spot having a plurality of spot properties,

generating scoring information for the spot properties of the at least one enrollment spot by generating probability distribution functions for at least one of true matches and false matches during an enrollment learning phase, and generating a family of scoring functions based upon the probability distribution functions for at least one of true matches and false matches,

generating at least one verification spot from finger biometric verification data, and

comparing the at least one enrollment spot with the at least one verification spot based upon the family of scoring functions to determine a match therewith.

38. A finger biometric sensor according to claim 37 wherein the spot properties are independent.

39. A finger biometric sensor according to claim 37 wherein the family of scoring functions are based upon a conditional false acceptance rate.

40. A finger biometric sensor according to claim 37 wherein the family of scoring functions are based upon a conditional likelihood ratio.

41. A finger biometric sensor according to claim 37 wherein said processor generates information content data from the finger biometric data for selecting the at least one enrollment spot based upon the information content data.