Computationally Efficient Feature Extraction and Matching Iris Recognition

• US 20100014718A1
• Filed: 04/17/2009
• Published: 01/21/2010
• Est. Priority Date: 04/17/2008
• Status: Active Grant

• Current Assignee: Biometricore, Inc.
• Original Assignee: Biometricore, Inc.
• Inventors: Savvides, Marios, Harun, Khalid, Li, Yunghui, Bhagavatula, Vijayakumar, Park, Sung Won
• Granted Patent: US 8,411,910 B2
• Time in Patent Office: Days
• US Class Current: 382/117
• CPC Class Codes: G06T 2207/10004 Still image; Photographic image
• CPC Class Codes: G06T 2207/30201 Face
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06V 40/172 Classification, e.g. identification
• CPC Class Codes: G06V 40/18 Eye characteristics, e.g. of the iris
• CPC Class Codes: G06V 40/19 Sensors therefor

Abstract

A method and system for uniquely identifying a subject based on an iris image. After obtaining the iris image, the method produces a filtered iris image by applying filters to the iris image to enhance discriminative features of the iris image. The method analyzes an intensity value for pixels in the filtered iris image to produce an iris code that uniquely identifies the subject. The method also creates a segmented iris image by detecting an inner and outer boundary for an iris region in the iris image, and remapping pixels in the iris region, represented in a Cartesian coordinate system, to pixels in the segmented iris image, represented in a log-polar coordinate system, by employing a logarithm representation process. The method also creates a one-dimensional iris string from the iris image by unfolding the iris region by employing a spiral sampling method to obtain sample pixels in the iris region, wherein the sample pixels are the one-dimensional iris string.

Claims

1. A method for uniquely identifying a subject based on an iris image of an eye of the subject, comprising:

obtaining the iris image;

applying at least one filter to the iris image to produce a filtered iris image, wherein said at least one filter enhances discriminative features of the iris image; and

applying at least one encoding process to the filtered iris image to produce the iris code, wherein said at least one encoding process analyzes an intensity value for pixels in the filtered iris image,

wherein the iris code uniquely identifies the subject.

2. The method of claim 1, wherein the iris image is a segmented iris image, having a two-dimensional format, and that is sampled from an original iris image employing a mapping function.

3. The method of claim 1, wherein the applying of said at least one filter further comprises:

applying at least one interpolation method to the filtered iris image to produce a resized iris image, wherein said at least one interpolation method interpolates an intensity value for pixels in the filtered iris image to downsample the filtered iris image; and

applying at least one differencing process to the resized iris image to produce at least one local difference iris image, wherein said at least one differencing process compares an intensity value for pixels in the resized iris image.

4. The method of claim 3, wherein said at least one differencing process compares the intensity value for pixels in the resized iris image that are adjacent neighbors.

5. The method of claim 3, wherein said at least one differencing process compares the intensity value for pixels in the resized iris image that are non-adjacent neighbors.

6. The method of claim 1, further comprising:

creating an iris mask for the iris code that determines indices of the iris code that are valid for matching purposes;

obtaining a second iris code and a second iris mask; and

computing a matching score by comparing the iris code and the iris mask to the second iris code and the second iris mask.

7. The method of claim 6, wherein the computing of the matching score further comprises:

creating a combined iris mask by combining the iris mask with the second iris mask;

computing a number of valid bits in the combined iris mask;

computing a number of bits that match in the iris code and the second iris code from only the indices denoted as valid in the combined iris mask; and

computing the matching score by dividing the number of matched bits by the number of valid bits.

8. A system for using a computer to uniquely identify a subject based on an iris image of an eye of the subject, comprising:

a memory device resident in the computer; and

a processor disposed in communication with the memory device, the processor configured to;

obtain the iris image;

apply at least one filter to the iris image to produce a filtered iris image, wherein said at least one filter enhances discriminative features of the iris image;

apply at least one encoding process to the filtered iris image to produce the iris code, wherein said at least one encoding process analyzes an intensity value for pixels in the filtered iris image,

wherein the iris code uniquely identifies the subject.

9. The system of claim 8, wherein the iris image is a segmented iris image, having a two-dimensional format, and that is sampled from an original iris image employing a mapping function.

10. The system of claim 8, wherein to apply said at least one filter, the processor is further configured to:

apply at least one interpolation method to the filtered iris image to produce a resized iris image, wherein said at least one interpolation method interpolates an intensity value for pixels in the filtered iris image to downsample the filtered iris image; and

apply at least one differencing process to the resized iris image to produce at least one local difference iris image, wherein said at least one differencing process compares an intensity value for pixels in the resized iris image.

11. The system of claim 10, wherein said at least one differencing process compares the intensity value for pixels in the resized iris image that are adjacent neighbors.

12. The system of claim 10, wherein said at least one differencing process compares the intensity value for pixels in the resized iris image that are non-adjacent neighbors.

13. The system of claim 8, wherein the processor further configured to:

create an iris mask for the iris code that determines indices of the iris code that are valid for matching purposes;

obtain a second iris code and a second iris mask; and

compute a matching score by comparing the iris code and the iris mask to the second iris code and the second iris mask.

14. The system of claim 13, wherein to compute the matching score, the processor is further configured to:

create a combined iris mask by combining the iris mask with the second iris mask;

compute a number of valid bits in the combined iris mask;

compute a number of bits that match in the iris code and the second iris code from only the indices denoted as valid in the combined iris mask; and

compute the matching score by dividing the number of matched bits by the number of valid bits.

15. A method for creating a segmented iris image from an image of an eye of a subject, comprising:

obtaining the image of the eye;

detecting an iris region in the image of the eye, the iris region having an inner boundary and an outer boundary; and

remapping pixels in the iris region, represented in a Cartesian coordinate system, to pixels in the segmented iris image, represented in a log-polar coordinate system, by employing a logarithm representation process.

16. The method of claim 15, wherein the detecting of the iris region further comprises:

detecting a first approximate circle to define the inner boundary of the iris region, wherein the inner boundary is between a pupil of the eye and the iris;

detecting a second approximate circle to define the outer boundary of the iris region, wherein the outer boundary is between the iris and a sclera of the eye;

applying an iris boundary fitting process to a number of initial points on the first approximate circle to refine the inner boundary of the iris region; and

applying the iris boundary fitting process to a number of initial points on the second approximate circle to refine the outer boundary of the iris region.

17. The method of claim 15, wherein the logarithm representation process further comprises:

calculating an intensity value for each pixel (h, α

) in the segmented iris image from the corresponding pixel (x, y) in the iris region, where x=(r+r_in)cos α

+x₀ and y=(r+r_in)sin α

+y₀, wherein each pixel in the iris image is located a distance r from the inner boundary, α

is a horizontal index of the corresponding pixel in the segmented iris image, r=s exp(β

h), where s is a scaling factor to map vertical indices in the segmented iris image linearly into [0, r_out−

r_in], where r_in is a distance from a center point in a pupil region of the eye to the inner boundary, and r_out is a distance from the center point in the pupil region of the eye to the outer boundary, and β

is a factor that varies the emphasis placed on the iris region near the inner boundary.

18. The method of claim 15, wherein the logarithm representation process further comprises:

calculating an intensity value for each pixel (h, α

) in the segmented iris image from the corresponding pixel (x, y) in the iris region, where x=(r+r_in)cos α

+x₀ and y=(r+r_in)sin α

+y₀, wherein each pixel in the iris image is located a distance r from the inner boundary, α

is a horizontal index of the corresponding pixel in the segmented iris image, r=s log_b(β

h), where s is a scaling factor to map vertical indices in the segmented iris image linearly into [0, r_out−

r_in], where r_in is a distance from a center point in a pupil region of the eye to the inner boundary, and r_out is a distance from the center point in the pupil region of the eye to the outer boundary, and β

is a factor that varies the emphasis placed on the iris region near the inner boundary.

19. A system for using a computer to create a segmented iris image from an image of an eye of a subject, comprising:

a memory device resident in the computer; and

a processor disposed in communication with the memory device, the processor configured to;

obtain the image of the eye;

detect an iris region in the image of the eye, the iris region having an inner boundary and an outer boundary; and

remapping pixels in the iris region, represented in a Cartesian coordinate system, to pixels in a segmented iris image, represented in a log-polar coordinate system, by employing a logarithm representation process.

20. The system of claim 19, wherein to detect the iris region, the processor is further configured to:

detect a first approximate circle to define the inner boundary of the iris region, wherein the inner boundary is between a pupil of the eye and the iris;

detect a second approximate circle to define the outer boundary of the iris region, wherein the outer boundary is between the iris and a sclera of the eye;

apply an iris boundary fitting process to a number of initial points on the first approximate circle to refine the inner boundary of the iris region; and

apply the iris boundary fitting process to a number of initial points on the second approximate circle to refine the outer boundary of the iris region.

21. The system of claim 19, wherein to employ the logarithm representation process, the processor is further configured to:

calculate an intensity value for each pixel (h, α

) in the segmented iris image from the corresponding pixel (x, y) in the iris region, where x=(r+r_in)cos α

+x₀ and y=(r+r_in)sin α

+y₀, wherein each pixel in the iris image is located a distance r from the inner boundary, α

is a horizontal index of the corresponding pixel in the segmented iris image, r=s exp(β

h), where s is a scaling factor to map vertical indices in the segmented iris image linearly into [0, r_out−

r_in], where r_in is a distance from a center point in a pupil region of the eye to the inner boundary, and r_out is a distance from the center point in the pupil region of the eye to the outer boundary, and β

is a factor that varies the emphasis placed on the iris region near the inner boundary.

22. The system of claim 19, wherein to employ the logarithm representation process, the processor is further configured to:

calculate an intensity value for each pixel (h, α

) in the segmented iris image from the corresponding pixel (x, y) in the iris region, where x=(r+r_in)cos α

+x₀ and y=(r+r_in)sin α

+y₀, wherein each pixel in the iris image is located a distance r from the inner boundary, α

is a horizontal index of the corresponding pixel in the segmented iris image, r=s log_b(β

h), where s is a scaling factor to map vertical indices in the segmented iris image linearly into [0, r_out−

r_in], where r_in is a distance from a center point in a pupil region of the eye to the inner boundary, and r_out is a distance from the center point in the pupil region of the eye to the outer boundary, and β

is a factor that varies the emphasis placed on the iris region near the inner boundary.

23. A method for creating a one-dimensional iris string from an image of an eye of a subject, comprising:

obtaining the image of the eye;

detecting an iris region in the image of the eye, the iris region having an inner boundary and an outer boundary; and

unfolding the iris region by employing a spiral sampling method to obtain sample pixels in the iris region,

wherein the sample pixels are the one-dimensional iris string.

24. The method of claim 23, wherein the detecting of the iris region further comprises:

detecting a first approximate circle to define the inner boundary of the iris region, wherein the inner boundary is between a pupil of the eye and the iris;

detecting a second approximate circle to define the outer boundary of the iris region, wherein the outer boundary is between the iris and a sclera of the eye;

applying an iris boundary fitting process to a number of initial points on the first approximate circle to refine the inner boundary of the iris region; and

applying the iris boundary fitting process to a number of initial points on the second approximate circle to refine the outer boundary of the iris region.

25. The method of claim 23, wherein the spiral sampling is outward from the inner boundary toward the outer boundary.

26. The method of claim 23, wherein the spiral sampling is inward from the outer boundary toward the inner boundary.

27. The method of claim 23, wherein the unfolding of the iris region further comprises:

normalizing the spiral sampling method to obtain a consistent number of sample pixels between the inner boundary and the outer boundary at every angle θ

.

28. A system for using a computer to create a one-dimensional iris string from an image of an eye of a subject, comprising:

a memory device resident in the computer; and

a processor disposed in communication with the memory device, the processor configured to;

obtain the image of the eye;

detect an iris region in the image of the eye, the iris region having an inner boundary and an outer boundary; and

unfold the iris region by employing a spiral sampling method to obtain sample pixels in the iris region,

wherein the sample pixels are the one-dimensional iris string.

29. The system of claim 28, wherein to detect the iris region, the processor is further configured to:

detect a first approximate circle to define the inner boundary of the iris region, wherein the inner boundary is between a pupil of the eye and the iris;

detect a second approximate circle to define the outer boundary of the iris region, wherein the outer boundary is between the iris and a sclera of the eye;

apply an iris boundary fitting process to a number of initial points on the first approximate circle to refine the inner boundary of the iris region; and

apply the iris boundary fitting process to a number of initial points on the second approximate circle to refine the outer boundary of the iris region.

30. The system of claim 28, wherein the spiral sampling is outward from the inner boundary toward the outer boundary.

31. The system of claim 28, wherein the spiral sampling is outward from the inner boundary toward the outer boundary.

32. The system of claim 28, wherein to unfold the iris region, the processor is further configured to:

normalize the spiral sampling method to obtain a consistent number of sample pixels between the inner boundary and the outer boundary at every angle θ

.