System and method for issuing a security unit after determining eligibility by image recognition

1. A method for issuing a security unit to an individual after determining eligibility by image recognition, the method comprising

providing an image of the individual;

finding a match between a representation of the image and a reference in a database associated with a reference individual by performing the steps of:

associating the image with a first vector;

projecting the first vector to form a second vector; and

evaluating a function of the second vector to match the image to the reference in the database; and

issuing a security unit to the individual provided the individual is eligible to receive said security unit.

2. The method of claim 1 , further comprising, before the step of providing, acquiring the image of the individual with an image acquisition device.

3. The method of claim 1 , further comprising providing a name of the individual.

4. The method of claim 1 , wherein the security unit includes at least one of a key, a card, and a photo ID.

5. The method of claim 3 , wherein the step of issuing includes the steps of:

determining whether the name of the individual does not match a name of the reference individual;

determining whether the individual has previously received a previous security unit; and

determining whether the reference individual is not allowed to receive the security unit, wherein an affirmative determination to any of the three determining steps deems the individual ineligible to receive said security unit and said security unit is not issued.

6. The method of claim 1 , wherein the first vector is substantially equal to Γ−Ψ, where Γ has components corresponding to intensities at pixels in the image, and Ψ is an average vector of a training set of vectors.

7. The method of claim 6 , wherein the step of projecting the first vector to form a second vector, Ω, includes

calculating a covariance matrix utilizing the training set of vectors and Ψ;

finding a set of eigenvectors associated with non-zero eigenvalues of the covariance matrix; and

forming the second vector by projecting the first vector onto a vector space spanned by the set of eigenvectors, the second vector being a vector representation, with respect to the set of eigenvectors, of the projection of the first vector, Ω=[P u (Γ−Ψ)] u .

8. The method of claim 1 , wherein the step of finding a match includes

associating the image with a first vector;

projecting the first vector to form a second vector;

to account for intra-object variation in the image, forming a third vector from the second vector and at least one of a first plurality of vectors representing the individual and a second plurality of vectors representing the reference individual; and

evaluating a function of the third vector to match the image to the reference in the database.

9. The method of claim 8 , wherein the first vector is substantially equal to Γ−Ψ, where Γ has components corresponding to intensities at pixels in the image, and Ψ is an average vector of a training set of vectors.

10. The method of claim 9 , wherein the step of projecting the first vector to form a second vector, Ω, includes

calculating a covariance matrix utilizing the training set of vectors and Ψ;

finding a set of eigenvectors associated with non-zero eigenvalues of the covariance matrix; and

forming the second vector by projecting the first vector onto a vector space spanned by the set of eigenvectors, the second vector being a vector representation, with respect to the set of eigenvectors, of the projection of the first vector, Ω=[P u (Γ−Ψ)] u .

11. The method of claim 10 , wherein the step of forming a third vector includes

calculating a second covariance matrix utilizing the first plurality of vectors; and

finding a second set of eigenvectors associated with non-zero eigenvalues of the second covariance matrix.

12. A system for issuing a security unit to an individual after determining eligibility by image recognition, the system comprising

a recognition system for finding a match between a representation of an image of the individual and a reference in a database associated with a reference individual, the recognition system including,

an image data preprocessor for associating an image of the individual with a first vector;

a projection module for projecting the first vector to form a second vector; and

a discrimination stage for evaluating a function of the second vector to match the representation of the image to the reference in the database; and

an issuing module for issuing a security unit to the individual provided the individual is eligible to receive said security unit.

13. The system of claim 12 , further comprising an image acquisition device for acquiring the image of the individual.

14. The system of claim 12 , wherein the issuing module is provided with a name of the individual, and the issuing module includes a determination module for determining whether

the name of the individual does not match a name of the reference individual;

the individual has previously received a previous security unit; and

the reference individual is not allowed to receive the security unit, wherein the system determines the individual is ineligible to receive the security unit if any of the three previous determinations are affirmative and the security unit is not issued to the individual.

15. The system of claim 12 , wherein the security unit includes at least one of a key, a card, and a photo ID.

16. The system of claim 12 , wherein the first vector is substantially equal to Γ−Ψ, where Γ has components corresponding to intensities at pixels in the image, and Ψ is an average vector of a training set of vectors.

17. The system of claim 16 , wherein the projection module

calculates a covariance matrix utilizing the training set of vectors and Ψ;

finds a set of eigenvectors associated with non-zero eigenvalues of the covariance matrix; and

forms the second vector by projecting the first vector onto a vector space spanned by the set of eigenvectors, the second vector being a vector representation, with respect to the set of eigenvectors, of the projection of the first vector, Ω=[P u (Γ−Ψ)] u .

18. The system of claim 12 , wherein the recognition system includes

an image data preprocessor for associating the image of the individual with a first vector;

a projection module for projecting the first vector to form a second vector;

to account for intrapersonal variation in the image, an extractor for forming a third vector from the second vector and at least one of a first plurality of vectors representing the individual and a second plurality of vectors representing the reference individual; and

a discrimination stage for evaluating a function of the third vector to match the representation of the image to the reference in the database.

19. The system of claim 18 , wherein the first vector is substantially equal to Γ−Ψ, where Γ has components corresponding to intensities at pixels in the image, and Ψ is an average vector of a training set of vectors.

20. The system of claim 19 , wherein the projection module

calculates a covariance matrix utilizing the training set of vectors and Ψ;

finds a set of eigenvectors associated with non-zero eigenvalues of the covariance matrix; and

forms the second vector by projecting the first vector onto a vector space spanned by the set of eigenvectors, the second vector being a vector representation, with respect to the set of eigenvectors, of the projection of the first vector, Ω=[P u (Γ−Ψ)] u .

21. The system of claim 18 , wherein the extractor

calculates a second covariance matrix utilizing the first plurality of vectors; and

finds a second set of eigenvectors associated with non-zero eigenvalues of the second covariance matrix.