Fingerprint symmetry axis extraction method and fingerprint recognition method using fingerprint symmetry axis

• US 7,471,812 B2
• Filed: 03/21/2005
• Issued: 12/30/2008
• Est. Priority Date: 03/22/2004
• Status: Expired due to Fees

• Current Assignee: LG Electronics, Inc. (LG Corporation)
• Original Assignee: LG Electronics, Inc. (LG Corporation)
• Inventors: Kim, Taek Soo, Cho, Sang Hyun
• Primary Examiner(s): Bella; Matthew C.
• Assistant Examiner(s): PERUNGAVOOR, SATHYANARAYA V
• Application Number:
  US11/086,872
• Publication Number:
  US 20050207625A1
• Time in Patent Office: 1,380 Days
• Field of Search: 382/125
• US Class Current: 382/125
• CPC Class Codes: G06V 40/1347 Preprocessing; Feature extraction

Abstract

A fingerprint symmetry axis extraction method is provided. The method includes detecting a core point of a fingerprint, setting imaginary symmetry axes intersecting the core point, and extracting an actual fingerprint symmetry axis from the imaginary symmetry axes. The actual fingerprint symmetry axis is one that can provide the largest symmetry measure being obtained between ridge range directions in paired blocks facing each other.

Claims

1. A fingerprint symmetry axis extraction method comprising:

detecting a core point of a fingerprint;

setting imaginary symmetry axes intersecting the core point; and

extracting an actual fingerprint symmetry axis from the imaginary symmetry axes, the actual fingerprint symmetry axis providing the largest symmetry measure between ridge range directions in paired blocks along the symmetry axis facing each other.

2. The fingerprint symmetry axis extraction method according to claim 1, wherein the blocks are contained in a predetermined circle area about the core point.

3. The fingerprint symmetry axis extraction method according to claim 1, wherein the imaginary symmetry axes are angularly spaced by an equal angle.

4. The fingerprint symmetry axis extraction method according to claim 1, further comprising extracting a more accurate fingerprint symmetry axis while thinly moving the actual symmetry axis by a minute angle.

5. The fingerprint symmetry axis extraction method according to claim 1, wherein the imaginary axes are formed within a range of 0−

π

.

6. The fingerprint symmetry axis extraction method according to claim 1, wherein the symmetry measure of the paired facing blocks is largest when a sum of angles defined by the ridge range directions in the paired facing blocks is π

.

7. The fingerprint symmetry axis extraction method according to claim 1, wherein the symmetry measure is determined for all of the blocks identically distant from the imaginary symmetry axes.

8. The fingerprint symmetry axis extraction method according to claim 1, wherein the symmetry measure is determined as a single value that is the sum of symmetry measures of the facing blocks.

9. The fingerprint symmetry axis extraction method according to claim 1, wherein the block is 16*16-pixel in a size.

10. A fingerprint recognition method comprising:

scanning a fingerprint image;

extracting a fingerprint symmetry axis by which a symmetry measured between ridge range directions along the symmetry axis of the scanned fingerprint image becomes maximum; and

performing an authentication for the scanned fingerprint image with reference to the extracted fingerprint symmetry axis.

11. The fingerprint recognition method according to claim 10, wherein the fingerprint symmetry axis becomes a reference for compensating rotation of the fingerprint image when the fingerprint image is inputted in a state where it is rotated.

12. The fingerprint recognition method according to claim 10, wherein an acceptance or rejection of the scanned fingerprint is determined whether feature vectors and the fingerprint symmetry axis are identical or not.

13. The fingerprint recognition method according to claim 10, wherein the number of the fingerprint symmetry axis is 1 for each fingerprint.

14. The fingerprint recognition method according to claim 10, wherein the authentication is performed by comparing feature vectors with each other after the symmetry axis of the fingerprint is co-aligned.

15. A fingerprint recognition method comprising:

calculating first and second sums, the first sum being obtained by summing angles of ridge range direction of paired blocks facing each other with reference to first imaginary symmetry axes, the second sum being obtained by adding an absolute value that is obtained by subtracting the first sum from a predetermined threshold value with respect to all of the blocks; and

performing an authentication by selecting one of the first imaginary symmetry axes, which has the smallest second sum.

16. The fingerprint recognition method according to claim 15, wherein the predetermined threshold values is π

.

17. The fingerprint recognition method according to claim 15, wherein the first sum is less than π

.

18. The fingerprint recognition method according to claim 15 wherein the blocks are arranged within a circle area having a predetermined radius from a core point of a fingerprint image.

19. The fingerprint recognition method according to claim 15, wherein the first imaginary symmetry axes are arranged intersecting a core point of a fingerprint image and angularly distant from each other by a predetermined first angle with reference to the core point.

20. The fingerprint recognition method according to claim 19, wherein second imaginary symmetry axes are further arranged intersecting the core point and distant from each other by a predetermined second angle after the first imaginary symmetry axes are arranged, the second angle being less than the first angle.