System and method for transforming fingerprints to improve recognition

• US 6,466,686 B2
• Filed: 01/07/1998
• Issued: 10/15/2002
• Est. Priority Date: 01/07/1998
• Status: Expired due to Term

• Current Assignee: Uniloc 2017 LLC (FIG LLC (d/b/a Fortress Investment Group LLC))
• Original Assignee: International Business Machines Corporation
• Inventors: Senior, Andrew William
• Primary Examiner(s): Boudreau, Leo
• Assistant Examiner(s): CHAWAN, SHEELA C
• Application Number:
  US09/003,988
• Publication Number:
  US 20020126883A1
• Time in Patent Office: 1,742 Days
• Field of Search: 382/116, 382/124, 382/125, 382/206, 382/209, 382/173, 382/190, 382/268, 382/272, 382/276, 382/285, 356/71, 348/875.34
• US Class Current: 382/124
• CPC Class Codes: G06V 40/1347 Preprocessing; Feature extraction

Abstract

A system and method reduces the amount of variation in a fingerprint image due to distortion from elastic deformation of the finger surface and projection onto a planar coordinate system. The amount and extent of distortion is estimated, and the process is inverted in a systematic way, so that the amount of distortion is minimized, and the distortion present is consistent, so that different images of the same print will have very similar distortion and thus matching of the fingerprints will be reliable and unaffected by the distortion.

Claims

1. A computer implemented method for estimating distortion present in a fingerprint, said distortion being variation in a fingerprint image due to elastic deformation of the finger surface and projection onto a planar coordinate system, comprising the steps of:

converting an input image of a fingerprint into a representation of locations of fingerprint ridges or valleys;

estimating distances between ridges or valleys locally throughout at least a selected portion of the representation;

combining the estimates of the distances to generate an estimated average spacing of ridges or valleys of the fingerprint; and

combining the average ridge spacing estimate with one or more of the local ridge spacing estimates to derive a measure of the elastic distortion of the fingerprint at one or more points on the fingerprint,

wherein said measure is used to reduce said distortion of said fingerprint.

2. The computer implemented method recited in claim 1 further comprising the step of capturing an image of a fingerprint and inputting the captured image to a computer on which the method is implemented, the step of capturing distorting the image of the fingerprint due to elastic deformation.

3. The computer implemented method recited in claim 1 further comprising the step of remapping the captured input image of a fingerprint using the estimated average spacing of ridges or valleys of the fingerprint to produce a substantially undistorted image of the fingerprint.

4. A method for reducing the amount of distortion present in a fingerprint representation, said distortion being variation in a fingerprint image due to elastic deformation of the finger surface and projection onto a planar coordinate system, comprising the steps of:

capturing a fingerprint image, the captured fingerprint image being distorted by the capture process;

preprocessing the captured fingerprint image and extracting information of the locations of ridges or valleys in the captured fingerprint image;

estimating ridge or valley spacing locally throughout at least a selected portion of the captured fingerprint image;

estimating an average ridge or valley spacing in the captured fingerprint image based on the estimate of ridge or valley spacing throughout the image;

estimating localized warps in the captured fingerprint image to normalize ridge or valley spacing by combining the average ridge spacing estimate with one or more of the local ridge spacing estimates;

extracting minutiae information from the captured fingerprint image,

applying local warps to the estimated minutiae locations to remove distortion; and

applying existing minutiae matching algorithms to establish fingerprint identity.

5. A system for reducing distortion present in a fingerprint representation, said distortion being variation in a fingerprint image due to elastic deformation of the finger surface and projection onto a planar coordinate system, comprising:

a computer for receiving representations of fingerprint images; and

a process implemented on a computer for estimating distortion of a fingerprint image by combining an average ridge spacing estimate with one or more local ridge spacing estimates and applying said estimate of distortion of the fingerprint image to remove the distortion by remapping the fingerprint image.

6. The system of claim 5 further comprising an input device connected to the computer and wherein a representation of a fingerprint image is an original fingerprint image captured by the input device.

7. The system of claim 6 wherein the input device is a scanner.

8. The system of claim 6 wherein the input device is an imaging subsystem.

9. The system of claim 5 further comprising a storage device connected to the computer and wherein a representation of a fingerprint image is a fingerprint image stored on the storage device.

10. The system of claim 5 further comprising a network connection to the computer and wherein a representation of a fingerprint image is transmitted to the computer from a remote location.

11. The system of claim 10 further comprising a second computer connected to the network connection at the remote location, the second computer transmitting the representation of a fingerprint image.

12. The system of claim 11 further comprising an input device connected to the second computer and wherein a representation of a fingerprint image is an original fingerprint image captured by the input device.

13. The system of claim 11 further comprising a storage device connected to the second computer and wherein a representation of a fingerprint image is a fingerprint image stored on the storage device.

14. The system of claim 5 further comprising a query process implemented on the computer to compare a remapped fingerprint image with one or more stored fingerprint representations for the purposes of finding a match.

15. The system of claim 14 wherein the stored fingerprint representations have been remapped from representations of original fingerprint images.

16. The system of claim 5 wherein the process implemented on the computer for estimating distortion of a fingerprint image estimates distances between pairs of ridges or valleys of a fingerprint image and combines estimates to give a global estimate of ridge or valley spacing and local estimates of ridge or valley spacing.

17. A system for reducing distortion present in a fingerprint representation, said distortion being variation in a fingerprint image due to elastic deformation of the finger surface and projection onto a planar coordinate system, comprising:

a computer for receiving representations of fingerprint images; and

a process implemented on a computer for estimating distortion of a fingerprint image by combining an average ridge spacing estimate with one or more local ridge spacing estimates and applying said estimate of distortion of the fingerprint image to remove the distortion by remapping the fingerprint image,

wherein said local ridge spacing estimates are resolved into horizontal and vertical components.

18. A system as in claim 17, wherein said resolution into horizontal and vertical components is derived by modeling said distortion as an anisotropic scaling, said scaling varying as a function of a position x, the separation between two ridges at any point having a true length ρ

, but said true length having been scaled by the distortion process so that the ratio of an estimated length ρ

′

to the local estimates, r′

(x), is resolved into orthogonal components $\frac{{\mathrm{\rho <br><br>}}^{\mathrm{\prime <br><br>}}}{r^{\mathrm{\prime <br><br>}}<br><br>\left(x\right)}=\frac{{\mathrm{\rho <br><br>}}^{\mathrm{\prime <br><br>}}}{r_x^{\mathrm{\prime <br><br>}}<br><br>\left(x\right)}<br><br>\text{\hspace{1em}<br><br>}<br><br>\hat{i}+\frac{{\mathrm{\rho <br><br>}}^{\mathrm{\prime <br><br>}}}{r_y^{\mathrm{\prime <br><br>}}<br><br>\left(x\right)}<br><br>\text{\hspace{1em}<br><br>}<br><br>\hat{j}$

where î and

ĵ

are the unit vectors in the x and y directions respectively.