Multimodal ocular biometric system and methods

• US 8,170,293 B2
• Filed: 09/10/2007
• Issued: 05/01/2012
• Est. Priority Date: 09/15/2006
• Status: Active Grant

• Current Assignee: Idemia Identity & Security USA, LLC (Advent International Corporation)
• Original Assignee: Identix, Inc. (Advent International Corporation)
• Inventors: Tosa, Yasunari, Usher, David, Ruzhitsky, Vladimir, Bashir, Faisal, Accomando, Nichoas A.
• Primary Examiner(s): Bhatnagar, Anand
• Assistant Examiner(s): PARK, SOO JIN
• Application Number:
  US11/898,190
• Publication Number:
  US 20080253622A1
• Time in Patent Office: 1,695 Days
• Field of Search: 382/115, 382/116, 382/118, 382/209
• US Class Current: 382/117
• CPC Class Codes: G06T 2207/30201 Face
• CPC Class Codes: G06T 7/12 Edge-based segmentation
• CPC Class Codes: G06T 7/149 involving deformable models, e.g. active contour models
• CPC Class Codes: G06T 7/168 involving transform domain methods
• CPC Class Codes: G06V 40/19 Sensors therefor

Abstract

Biometric systems capture and combine biometric information from more than one modality, employing digital processing algorithms to process and evaluate captured images having data for a biometric characteristic. Such digital algorithms may include a pupil segmentation algorithm for determining a pupil image in the captured image, an iris segmentation algorithm for determining an iris image in the captured image, an eyelid/eyelash segmentation algorithm for determining an eyelid/eyelash image in the captured image, and an algorithm for measuring the focus on the iris. Some embodiments employ an auto-capture process which employs such algorithms, in part, to evaluate captured images and obtain the best possible images for biometric identification.

Claims

1. A method for segmenting a biometric characteristic from an eye image captured by a biometric device, the method comprising the steps of:

defining one or more first search windows in a captured eye image;

identifying a first set of points corresponding to peaks in an image intensity gradient in each of the one or more first search windows;

determining, from the first set of points, segments according to image intensity transitions, each segment having a center point;

determining a position for the center point of each segment;

grouping sets of the segments into a first set of clusters according to an analysis of the positions of the center points for the segments; and

selecting, from the first set of clusters, a second set of clusters corresponding to a pupil image;

fitting a first model template to the points for each set of segments corresponding to the second set of clusters, the fitted first model template corresponding to an initial pupil outer boundary for the pupil image;

defining one or more second search windows in relation to the initial pupil outer boundary;

identifying a second set of points corresponding to peaks in an image intensity gradient in each of the one or more second search windows; and

fitting a second model template to the second set of points, the fitted second model template corresponding to a refined pupil outer boundary for the pupil image.

2. The method according to claim 1, wherein the first model template and the second model template are circles.

3. The method according to claim 1, wherein the one or more first search windows are rectangular search windows extending horizontally across the eye image.

4. The method according to claim 1, wherein the one or more second search windows are rectangular search windows with longitudinal axes extending perpendicularly to the initial pupil outer boundary.

5. The method according to claim 1, wherein the image intensity transitions comprise a light-to-dark transition on a first side and a dark-to-light transition on a second side.

6. The method according to claim 1, wherein the step of selecting a second set of clusters comprises the steps of selecting, from the first set of clusters, corresponding to a shape feature.

7. The method according to claim 6, wherein the step of selecting a second set of clusters further comprises applying a weighting score to each cluster in the first set of clusters according to pupil-related scoring criteria and sorting the first set of clusters according to the weighting scores.

8. The method according claim 1, further comprising:

defining one or more third search windows in relation to the pupil image;

identifying a third set of points corresponding to image intensity transitions on a right side of the pupil image and to a left side of the pupil image, the third set of points corresponding to edge points;

fitting a third model template to the third set of points, the fitted third model template corresponding to an initial iris outer boundary for the iris image;

defining one or more fourth search windows in relation to the initial iris outer boundary for the iris image;

identifying a fourth set of points corresponding to peaks in an image intensity gradient in each of the one or more fourth search windows; and

fitting a fourth model template to the fourth set of points, the fitted fourth model template corresponding to a refined iris outer boundary for the iris image,

wherein the refined pupil outer boundary of the pupil image corresponds with an iris inner boundary of the iris image.

9. The method according to claim 8, wherein the third set of points comprises two edge points.

10. The method according to claim 8, wherein the third model template and the fourth model template are circles.

11. The method according to claim 8, wherein the one or more third search windows are rectangular search windows with longitudinal axes which start at a center of the pupil image and extend downwardly at an angle from a horizontal line of the eye image.

12. The method according to claim 8, wherein the one or more fourth search windows are rectangular search windows with a longitudinal axis extending perpendicularly to the initial iris outer boundary.

13. The method according to the claim 8, further comprising identifying an annular region corresponding to the iris image, the annular region being defined by the refined iris outer boundary for the iris image and the refined iris inner boundary for the iris image.

14. The method according to claim 13, further comprising computing an amount of occlusion of the iris image.