Reconstruction of biometric image templates using match scores

• US 8,165,352 B1
• Filed: 08/06/2008
• Issued: 04/24/2012
• Est. Priority Date: 08/06/2007
• Status: Active Grant

• Current Assignee: University of South Florida (State University System of Florida)
• Original Assignee: University of South Florida (State University System of Florida)
• Inventors: Sarkar, Sudeep, Mohanty, Pranab, Kasturi, Rangachar
• Primary Examiner(s): Allison, Andrae S
• Application Number:
  US12/187,028
• Time in Patent Office: 1,357 Days
• Field of Search: 382/115, 382/116, 382/117, 382/118, 382/305, 382/224, 382/226, 382/228, 340/5.53, 340/5.83
• US Class Current: 382/115
• CPC Class Codes: G06F 18/2137 based on criteria of topology preservation, e.g. multidimensional scaling or self-organising maps
• CPC Class Codes: G06V 10/7715 Feature extraction, e.g. by transforming the feature space, e.g. multi-dimensional scaling [MDS]; Mappings, e.g. subspace methods
• CPC Class Codes: G06V 40/172 Classification, e.g. identification
• CPC Class Codes: G06V 40/50 Maintenance of biometric data or enrolment thereof

Abstract

A method of reconstructing biometric face image templates of a face recognition system (FRS) using the match scores or distances provided by the FRS. The match scores represent the distance between a image introduced to the FRS and the unknown image template stored in the FRS. The present method uses an affine transformation approximating the unknown algorithm within the FRS and the match scores provided by the FRS to determine the coordinates of the unknown target template. The coordinates of the unknown target template are then applied to a pseudo-inversion of the affine transformation to produce a reconstructed image template of the unknown target. This reconstructed image template can then be used to ‘break-in’ to the FRS.

Claims

1. A method of reconstructing an unknown enrolled biometric target template of a recognition system, the method comprising:

inputting a plurality of known biometric templates into the recognition system;

calculating a match score for each of the plurality of known biometric templates and the unknown enrolled biometric target template of the recognition system;

converting the match score for each of the plurality of known biometric templates into a plurality of dot product distances;

determining a set of configuration points for the plurality of dot product distances that preserves the match score for each of the plurality of known biometric templates and the unknown enrolled biometric target template of the recognition system, wherein the set of configuration points comprises a set of vectors wherein a sum of the vectors in the set of vectors is equal to zero such that the center of the set of configuration points is at the origin;

providing an affine transformation approximating the recognition system based upon the set of configuration points;

identifying coordinates for each of the configuration points in the set of configuration points in the affine transformation;

approximating coordinates for the unknown enrolled biometric target template in the affine transformation; and

performing a pseudo-inversion of the affine transformation to reconstruct the unknown enrolled biometric target template.

2. The method of claim 1, wherein the plurality of known biometric templates are distinct and independent.

3. The method of claim 1, wherein determining a set of configuration points further comprising employing multidimensional scaling to determine the set of configuration points.

4. A method of reconstructing an unknown enrolled face image target template of a Face Recognition System, the method comprising

inputting a plurality of known face image templates into the Face Recognition System;

calculating a match score for each of the plurality of known face image templates and the unknown enrolled face image target template of the Face Recognition System;

converting the match score for each of the plurality of known face image templates into a plurality of dot product distances;

determining a set of configuration points for the plurality of dot product distances that reserves the match score for each of the plurality of known face image templates and the unknown enrolled face image target template of the Face Recognition System, wherein the set of configuration points comprises a set of vectors wherein a sum of the vectors in the set of vectors is equal to zero such that the center of the set of configuration points is at the origin;

providing an affine transformation approximating the Face Recognition system based upon the set of configuration points;

identifying coordinates for each of the configuration points in the set of configuration points in the affine transformation;

approximating coordinates for the unknown enrolled face image target template in the affine transformation; and

performing a pseudo-inversion of the affine transformation to reconstruct the unknown enrolled face image target template.

5. The method of claim 4, wherein determining a set of configuration points further comprising employing multidimensional scaling to determine the set of configuration points.

6. The method of claim 4, wherein the plurality of known face image templates are distinct and independent.

7. The method of claim 4, wherein the match scores are similarity match scores.

8. The method of claim 4, wherein the match scores are dissimilarity match scores.

9. The method of claim 4, wherein approximating coordinates for the unknown enrolled face image template further comprises embedding the unknown enrolled face image template into the affine transformation based on the match score for each of the plurality of known face image templates.