Zero reference based ridge flow map

• US 9,092,652 B2
• Filed: 03/15/2013
• Issued: 07/28/2015
• Est. Priority Date: 06/29/2012
• Status: Active Grant

• Current Assignee: Apple Inc.
• Original Assignee: Apple Inc.
• Inventors: Marciniak, Craig A., Westerman, Wayne C., Han, Byron B.
• Primary Examiner(s): Mehta, Bhavesh
• Assistant Examiner(s): Dunphy, David F
• Application Number:
  US13/831,578
• Publication Number:
  US 20140003680A1
• Time in Patent Office: 865 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06V 40/12 Fingerprints or palmprints
• CPC Class Codes: G06V 40/1359 Extracting features related to ridge properties; Determining the fingerprint type, e.g. whorl or loop

Abstract

A method includes receiving fingerprint image data at a fingerprint recognition sensor, where the fingerprint image data are associated with an authorized user. The fingerprint image data are transformed into a substantially rotationally invariant representation, which is maintained in a database of enrolled fingerprint information. Processed fingerprint image data from an accessing user are compared with the substantially rotationally invariant representation of the fingerprint image data from the authorized user.

Claims

1. A method comprising:

receiving fingerprint image data associated with an authorized user by a fingerprint recognition sensor;

transforming the fingerprint image data associated with the authorized user into a substantially rotationally invariant representation thereof, wherein transforming the fingerprint image data associated with the authorized user comprises determining a set of directions based on gradients defined at different points in the fingerprint image data, and determining a set of differences of the directions;

maintaining the substantially rotationally invariant representation of the fingerprint image data associated with the authorized user in a database of enrolled fingerprint information; and

comparing processed fingerprint image data from an accessing user with the substantially rotationally invariant representation of the fingerprint image data from the authorized user by comparing a first histogram for the set of differences based on the fingerprint image data associated with the authorized user and a second histogram for a corresponding set of differences based on the fingerprint image data from the accessing user.

2. The method of claim 1, wherein comparing the processed fingerprint image data is performed in response to the accessing user attempting to match fingerprints with the authorized user.

3. The method of claim 2, further comprising authorizing the accessing user based on a match between the substantially rotationally invariant representation of the fingerprint image data from the authorized user and the processed fingerprint image data from the accessing user, wherein the fingerprint image data associated with the authorized user and the processed fingerprint image data from the accessing user are obtained at substantially different angular orientations with respect to the fingerprint recognition sensor.

4. The method of claim 1, further comprising determining a landmark in the fingerprint image data associated with the authorized user, wherein the set of differences is substantially rotationally invariant about the landmark.

5. The method of claim 4, wherein the landmark is based on a set of Poincare index values determined from the set of directions, and wherein the Poincare index values are indicative of a feature of the fingerprint image data associated with the authorized user.

6. The method of claim 5, wherein comparing a first histogram for the set of differences based on the fingerprint image data associated with the authorized user comprises comparing a first histogram for the set of Poincare index values based on the fingerprint image data associated with the authorized user, and wherein comparing a second histogram for a corresponding set of differences based on the fingerprint image data from the accessing user comprises comparing a second histogram for a corresponding set of Poincare index values based on the fingerprint image data from the accessing user.

7. A fingerprint image processing system comprising:

a fingerprint sensor configured for receiving first fingerprint image data of an authorized user and second fingerprint image data of an accessing user, wherein the first and second fingerprint image data are oriented at substantially different angles with respect to the fingerprint sensor;

a database configured for storing a substantially rotationally invariant representation of the first fingerprint image data of the authorized user, wherein the rotationally invariant representation of the first fingerprint image data comprises a set of determined directions based on gradients defined at different points in the first fingerprint image data and a set of determined differences of the directions; and

a processor configured for authorizing the accessing user based on comparing the substantially rotationally invariant representation of the first fingerprint image data of the authorized user with a substantially rotationally invariant representation of the second fingerprint image data of the accessing user by comparing a first histogram for the set of differences based on the first fingerprint image data associated with the authorized user and a second histogram for a corresponding set of differences based on the fingerprint image data from the accessing user.

8. The fingerprint image processing system of claim 7, wherein the processor is configured to determine a landmark in the fingerprint image data associated with the authorized user, wherein the set of differences is substantially rotationally invariant about the landmark.

9. The fingerprint image processing system of claim 8, wherein the landmark is based on a set of Poincare index values determined from the set of directions, and wherein the Poincare index values are indicative of a feature of the fingerprint image data associated with the authorized user.

10. The fingerprint image processing system of claim 9, wherein the first histogram comprises a first histogram for the set of Poincare index values based on the fingerprint image data associated with the authorized user and the second histogram comprises a second histogram for a corresponding set of Poincare index values based on the fingerprint image data from the accessing user.

11. A non-volatile computer readable data storage medium having program data stored thereon, the program data executable by a computer processor to perform a method comprising:

receiving first fingerprint image data from an authorized user, the first fingerprint image data oriented at a first angle with respect to a fingerprint image sensor;

transforming the first fingerprint image data from the authorized user into a substantially rotationally invariant representation thereof;

storing the substantially rotationally invariant representation of the first fingerprint image data from the authorized user in a database;

receiving second fingerprint image data from an accessing user, the second fingerprint image data oriented at a second angle with respect to the fingerprint image sensor, where the second angle is different from the first angle;

transforming the second fingerprint image data from the accessing user into a substantially rotationally invariant representation thereof, wherein the substantially rotationally invariant transformations of the first and second fingerprint image data comprise first and second sets of Poincare index values determined from a set of directions that are based on gradients defined at different points in the first and second fingerprint image data;

comparing the substantially rotationally invariant transformations of the first and second fingerprint image data to determine a match by comparing a first histogram for the first set of Poincare index values and a second histogram for the second set of Poincare index values; and

authorizing the accessing user, based on the match.