METHOD AND APPARATUS FOR AUTHENTICATING AREA BIOMETRIC SCANNERS

• US 20120300988A1
• Filed: 05/24/2012
• Published: 11/29/2012
• Est. Priority Date: 07/19/2010
• Status: Active Grant

• Current Assignee: University of Maryland
• Original Assignee: University of Maryland
• Inventors: IVANOV, Vladimir Iankov, Baras, John S.
• Granted Patent: US 8,942,430 B2
• Time in Patent Office: Days
• US Class Current: 382/115
• CPC Class Codes: G06F 18/22 Matching criteria, e.g. proximity measures
• CPC Class Codes: G06T 5/00 Image enhancement or restoration
• CPC Class Codes: G06V 10/751 Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
• CPC Class Codes: G06V 40/12 Fingerprints or palmprints
• CPC Class Codes: G06V 40/13 Sensors therefor
• CPC Class Codes: G06V 40/1306 non-optical, e.g. ultrasonic or capacitive sensing
• CPC Class Codes: G06V 40/40 Spoof detection, e.g. liveness detection
• CPC Class Codes: G06V 40/70 Multimodal biometrics, e.g. combining information from different biometric modalities

Abstract

Methods and apparatuses for authenticating a biometric scanner, such as area type finger print scanners, involves estimating unique intrinsic characteristics of the scanner (scanner pattern), that are permanent over time, and can identify a scanner even among scanners of the same manufacturer and model. Image processing and analysis are used to extract a scanner pattern from images acquired with the scanner. The scanner pattern is used to verify whether the scanner that acquired a particular image is the same as the scanner that acquired one or several images during enrollment of the biometric information. Authenticating the scanner can prevent subsequent security attacks using counterfeit biometric information on the scanner, or on the user authentication system.

Claims

1. A method for authenticating scanners, including fingerprint area scanners, said method comprising:

using an electronic processing circuit configured to perform the following,

(a) acquiring at least one enrolled image for comparison to at least one query image subsequently input to said scanner;

(b) decomposing said at least one enrolled image and said at least one query image using wavelets to, respectively, compute enrolled wavelet coefficients and query wavelet coefficients;

(c) performing wavelet reconstruction by, respectively, setting to zero at least the LL-subband coefficients of said enrolled wavelet coefficients to compute enrolled residuals and at least the LL-subband coefficients of said query wavelet coefficients to compute query residuals;

(d) masking useful pixels of said enrolled residuals and said query residuals;

(e) computing a similarity score between common pixels of said useful pixels from said enrolled residuals and said query residuals; and

(f) comparing said similarity score with a threshold value to determine whether said at least one query image inputted to the scanner was acquired by the same scanner that acquired said at least one enrolled image.

2. The method of claim 1 wherein said at least one enrolled image and said at least one query image represent biometric information provided by two different individuals.

3. The method of claim 1 wherein said masking uses at least one predetermined criterion for determining said useful pixels.

4. The method of claim 1 wherein said wavelets are any one of biorthogonal wavelets, Daubechies wavelets, and symlets wavelets.

5. The method of claim 1 wherein said computing is any one of computing a correlation coefficient, computing a normalized correlation, and computing a relative mean square error.

6. The method of claim 1 further comprising any one of inverting the pixel values of said at least one enrolled image and the pixel values of said at least one query image after (a) and inverting the pixel values of said enrolled residuals and the pixel values of said query residuals after (c).

7. A method for identifying scanners, including fingerprint area scanners, said method comprising:

using an electronic processing circuit configured to perform the following,

(a) acquiring and processing at least one image;

(b) using wavelets of said at least one image to compute wavelet coefficients;

(c) performing wavelet reconstruction by setting to zero at least the LL-subband coefficients of said wavelet coefficients to compute residuals;

(d) masking useful pixels of said residuals; and

(e) computing a sequence of numbers from said residuals that contains information which represents the scanner that acquired said at least one image.

8. The method of claim 7 wherein said masking uses at least one predetermined criterion for determining said useful pixels.

9. The method of claim 7 wherein said wavelets are any one of biorthogonal wavelets, Daubechies wavelets, and symlets wavelets.

10. The method of claim 7 wherein said computing is any one of computing a correlation coefficient, computing a normalized correlation, and computing a relative mean square error.

11. The method of claim 7 further comprising any one of inverting the pixel values of said at least one enrolled image and the pixel values of said at least one query image after (a) and inverting the pixel values of said enrolled residuals and the pixel values of said query residuals after (c).

12. A system for authenticating scanners, including fingerprint area scanners, said system comprising:

acquiring means for acquiring at least one enrolled image for comparison to at least one query image subsequently input to said scanner;

processing means for decomposing said at least one enrolled image and said at least one query image using wavelets to, respectively, compute enrolled wavelet coefficients and query wavelet coefficients;

wavelet means for performing wavelet reconstruction by, respectively, setting to zero at least the LL-subband coefficients of said enrolled wavelet coefficients to compute enrolled residuals and at least the LL-subband coefficients of said query wavelet coefficients to compute query residuals;

masking means for masking useful pixels of said enrolled residuals and said query residuals;

computing means for computing a similarity score between common pixels of said useful pixels from said enrolled residuals and said query residuals; and

comparing means for comparing said similarity score with a threshold value to determine whether said at least one query image inputted to the scanner was acquired by the same scanner that acquired said at least one enrolled image.

13. The system of claim 12 wherein said at least one enrolled image and said at least one query image represent biometric information provided by two different individuals.

14. The system of claim 12 wherein said masking means uses at least one predetermined criterion for determining said useful pixels.

15. The system of claim 12 wherein said processing means uses wavelets of any one of biorthogonal wavelets, Daubechies wavelets, and symlets wavelets.

16. The system of claim 12 wherein said computing means computes any one of computing a correlation coefficient, computing a normalized correlation, and computing a relative mean square error.

17. The system of claim 12 further comprising any one of an inverting means for inverting the pixel values of said at least one enrolled image and said at least one query image and an inverting means for inverting the pixel values of said enrolled residuals and the pixel values of said query residuals.

18. A system for identifying scanners, including fingerprint area scanners, said system comprising:

acquiring means for acquiring and processing at least one image;

processing means for decomposing said at least one image using wavelets to compute wavelet coefficients;

processing means for performing wavelet reconstruction by setting to zero at least the LL-subband coefficients of said wavelet coefficients to compute residuals;

masking means for masking useful pixels of said residuals; and

computing means for computing a sequence of numbers from said residuals that contains information which represents the scanner that acquired said at least one image.

19. The system of claim 18 wherein said masking means uses at least one predetermined criterion for determining said useful pixels.

20. The system of claim 18 wherein said processing means uses wavelets of any one of biorthogonal wavelets, Daubechies wavelets, and symlets wavelets.

21. The system of claim 18 wherein said computing means computes any one of computing a correlation coefficient, computing a normalized correlation, and computing a relative mean square error.

22. The system of claim 18 further comprising any one of an inverting means for inverting the pixel values of said at least one image and an inverting means for inverting the pixel values of said residuals.