IRIS RECOGNITION SYSTEM HAVING IMAGE QUALITY METRICS

• US 20070274570A1
• Filed: 03/02/2007
• Published: 11/29/2007
• Est. Priority Date: 01/26/2005
• Status: Active Grant

• Current Assignee: Gentex Corporation
• Original Assignee: Honeywell International Inc.
• Inventors: Hamza, Rida
• Granted Patent: US 8,050,463 B2
• Time in Patent Office: Days
• US Class Current: 382/117
• CPC Class Codes: G06V 10/993 Evaluation of the quality of the acquired pattern
• CPC Class Codes: G06V 40/193 Preprocessing; Feature extraction

Abstract

An iris recognition system implementing image quality metrics to assess the quality of an acquired eye image for reliable operation. Images with low image quality may be rejected or flagged based upon the application. The image quality may be determined with a preprocessing module in the recognition system. The processing may be configured based on a quality assessment

Claims

1. An iris image quality metric system comprising:

an iris image source; and

a quality processing module connected to the image source; and

wherein the quality processing module is for providing a quality assessment of an iris image.

2. The system of claim 1, wherein the assessment of an iris image is provided according to one or more of the following image conditions which comprise:

blur;

defocus;

closure;

obscuration;

off-angle detection;

reflection; and

/or

excessive dilation.

3. The system of claim 1, wherein:

a quality assessment of the iris image is according to blur; and

blur is measured according to high frequency distortions from course to fine wavelet coefficients.

4. The system of claim 1, wherein a quality assessment of the iris image is based on an XOR-ing of codes of two patches of the iris image to measure a discrepancy among bits of the patches.

5. The system if claim 1, wherein a quality assessment of the iris image is based on measuring similarity of two patches of the iris image using intensity MSE, contrast similarity, intensity distribution similarity, peak signal to noise ratio, and/or correlation between intensities of the two patches

6. The system of claim 1, wherein a quality measurement relative to blur and/or defocus of the iris image is based on a measurement of energy at high frequencies of the spectral domain of the image.

7. The system of claim 6, wherein:

the measurement of energy at high frequencies of the spectral domain of the image is compared to a measurement of energy at high frequencies of the spectral domain of the image without blur and/or defocus; and

an amount that the energy at high frequencies of the spectral domain of the image is less than the energy at high frequencies of the spectral domain of the image without blur and/or defocus is proportional to the blur and/or defocus of the image.

8. The system of claim 3, wherein the high frequency distortions among the course to fine wavelet coefficients are indicated by a comparison of linear frequency distortion filter outputs at multiple stages of a dyadic decomposition to measure a discrepancy among the stages.

9. The system of claim 3, wherein high frequency content is compared at wavelet decompositions between two patches of the iris image for similarity purposes.

10. The system of claim 1, wherein a quality assessment of the iris image relative to eye closure is according to an inner border profile of the iris in the image.

11. The system of claim 10, wherein the inner border profile is estimated according to a measurement of an amount that a detected curve of the profile is similar to an elliptic-like shape, and a measurement of an amount of exposure of a map of the iris.

12. The system of claim 1, wherein a quality assessment of the iris image relative to iris obscuration is according to a computation of an integral of virtually total area between an eyelid curve and an iris inner boundary of an eye from which the iris image is taken.

13. The system of claim 1, wherein a quality assessment of the iris image relative to an amount of reflection by the iris in the image is according to an iris curve fitting and a contrast thresholding.

14. The system of claim 1, wherein a quality assessment of the iris image is according to a comparison of a location of edges of the pupil with a set of defined limits for normal operation of a pupil dilation.

15. An iris image preprocessing system comprising:

an iris image module;

an iris image quality measurement module connected to the iris image module; and

an evaluator connected to the iris image module.

16. The system of claim 15, further comprising:

a rehabilator connected to the evaluator; and

wherein;

the evaluator is for receiving a quality measurement of an iris image from the quality measurement module, and for indicating whether the iris image is acceptable or unacceptable for further processing; and

the evaluator is for indicating whether an iris image that is unacceptable should be rejected or be rehabilitated for further processing.

17. The system of claim 16, wherein further processing comprises segmentation of the iris image.

18. A method for assessing quality of an image comprising:

receiving an iris image;

measuring at least one quality of the iris image; and

evaluating the at least one quality to determine whether the iris image is adequate relative to the at least one quality for further processing.

19. The method of claim 18, further comprising determining whether an iris image, which is not adequate relative to the at least one quality for further processing, is adequate for rehabilitation.

20. The method of claim 18, further comprising:

measuring a first energy, at high frequencies, of the spectral domain of the iris image;

determining a second energy, at high frequencies, of the spectral domain of a model iris image which is adequate for further processing; and

wherein if the first energy is within a set percentage of the second energy, then the iris image is adequate, relative to the at least one quality, for further processing.

21. The method of claim 18, further comprising:

obtaining two different localized patches from an iris code of the iris image;

XOR-ing the two patches;

measuring an amount of discrepancy among cross-matched bits of the two patches;

determining whether the amount of discrepancy is greater than a set amount; and

wherein if the amount of discrepancy is greater than a set amount, then the iris image is adequate, relative to the at least one quality, for further processing.

22. The method of claim 18, further comprising:

measuring an amount of difference between the inner border profile in the iris image and an elliptic-like shape; and

measuring a percentage of the iris map of the iris image that is exposed; and

wherein;

if the amount of difference is less than a set difference for an iris image adequate for further processing relative to the at least one quality, then the image is adequate for further processing relative to the at least one quality; and

if the percentage is greater than a set percentage for an iris image adequate for further processing relative to the at least one quality, then the image is adequate for further processing relative to the at least one quality.

23. An iris image quality metric system comprising:

an iris image source; and

a quality processing module connected to the image source; and

wherein;

the quality processing module is for providing a quality assessment of an iris image; and

processing is configured based upon the quality assessment of the iris image.