SYSTEM FOR IRIS DETECTION TRACKING AND RECOGNITION AT A DISTANCE

• US 20100239119A1
• Filed: 05/09/2006
• Published: 09/23/2010
• Est. Priority Date: 03/03/2006
• Status: Active Grant

• Current Assignee: Gentex Corporation
• Original Assignee: Honeywell International Inc.
• Inventors: Bazakos, Michael E., Au, Kwong Wing, Kilgore, George A.
• Granted Patent: US 8,064,647 B2
• Time in Patent Office: Days
• US Class Current: 382/103
• CPC Class Codes: G06V 40/167 using comparisons between temporally consecutive images
• CPC Class Codes: G06V 40/18 Eye characteristics, e.g. of the iris
• CPC Class Codes: G06V 40/19 Sensors therefor
• CPC Class Codes: G06V 40/193 Preprocessing; Feature extraction
• CPC Class Codes: G06V 40/197 Matching; Classification

Abstract

A stand-off range or at-a-distance iris detection and tracking for iris recognition having a head/face/eye locator, a zoom-in iris capture mechanism and an iris recognition module. The system may obtain iris information of a subject with or without his or her knowledge or cooperation. This information may be sufficient for identification of the subject, verification of identity and/or storage in a database.

Claims

1-23. -23. (canceled)

24. A system for iris detection at a distance of non-cooperative or cooperative subjects, comprising:

a multi-band imaging camera for detecting skin of a subject;

a first mechanism for determining with specific algorithms whether the skin is of a region containing at least a part of a face of the subject containing at least one eye, by locating several facial features such as eyes, eyebrows, nose or mouth and their positions relative to each other;

an adjustable zoom camera, connected to a processor, for obtaining initial close-up high resolution images of the region that contains at least one eye, the eye not necessarily looking directly at the camera; and

a second mechanism, connected to the camera, for determining, with eye tracking algorithms, whether the initial images contain a best iris position; and

wherein;

the best iris position is determined with the algorithms by maximizing a function that depends on key features of the iris to provide a point of maximization;

wherein the processor is configured to send a reacquisition request to the zoom camera to obtain additional close-up high resolution images of the region that contains at least one eye if the initial images do not contain a best iris position;

wherein if the initial images do contain a best iris position, the zoom camera takes several secondary close-up images of each iris region at the point of maximization; and

wherein the secondary images are passed on to be processed by an iris recognition device.

25. The system of claim 24, wherein the imaging camera comprises an infrared band.

26. The system of claim 24, wherein the first mechanism is a face detection and tracking system.

27. The system of claim 24, wherein the iris recognition device is for identifying the subject upon recognition of an iris in the images of iris regions of the subject.

28. The system of claim 24, wherein the eye is not necessarily looking directly at the camera since the eye may be of a non-cooperative subject.

29. A system for iris detection, tracking and recognition of a non-cooperative or cooperative subject at a distance, comprising:

an acquisition module comprising;

a multi-band imaging camera for skin detection of a subject;

a face detection and tracking system for determining if detected skin is part of a face of the subject by locating several facial features, such as eyes, and extracting the facial features; and

wherein;

from one or more facial features, information containing face orientation with respect to line of sight and eye location is sent to a juncture for determining whether the information provides a sufficiently good view; and

metrics for a good view comprise face symmetry and orientation based on the facial features;

wherein the acquisition module includes metrics that signals the face detection and tracking system to re-initiate the facial features extraction when there is not a good view, the acquisition module repeating the facial feature extraction until a good view is achieved;

when there is a good view, the information containing face orientation and eye location is sent to a zoom-in iris capture mechanism;

the zoom-in iris capture mechanism comprises;

a zoom-in and framing mechanism for zooming in and framing an eye region containing an iris;

an iris segmentation mechanism connected to the zoom-in and framing mechanism; and

wherein the iris segmentation mechanism is for extracting features of the iris by approximating inner and outer borders of the iris by ellipses and performing one-dimensional segmentation of the iris in a polar domain.

30. The system of claim 29, further comprising a mechanism for converting features of the iris into an iris numeric code.

31. The system of claim 30, further comprising a mechanism for comparing the iris numeric code with previously generated iris numeric codes for verification and identification of the iris numeric code.

32. The system of claim 29, wherein the segmentation mechanism is further for characterizing shape and texture of a structure of the iris having interlacing blocks of freckles, coronas, furrows, crypts and stripes.

33. The system of claim 32, wherein mixture modeling is used to handle variation in the texture of the structure of the iris.

34. The system of claim 29, wherein the inner and outer borders of the iris are approximated by ellipses using snake delineation.

35. The system of claim 29, wherein the outer border of the iris is instead approximated with irregular edges due to eyelids and eyelashes.

36. The system of claim 29, wherein one band of the camera is infrared.

37. A method for detecting an iris at a distance of a non-cooperative or cooperative subject, comprising:

providing a system for iris detection, the system including an acquisition module, a processor, and at least one camera, the method including using the system to perform the following steps;

scanning for a subject;

detecting skin of the subject;

determining whether the skin is of a region containing a face of the subject by locating one or more facial features such as eyes;

obtaining a plurality detailed images of the region containing the face and having at least one eye;

determining if one of the plurality of images includes an image of an iris of the at least one eye that shows a best position or view of the iris based on a maximizing a function that depends on features of the iris;

if no image shows a best position or view of the iris, the system sends a reacquisition signal to the acquisition module and the method steps are repeated until a successful image of the iris showing a best position of view of the iris is achieved; and

when a successful image of the iris is achieved, extracting features of the iris by determining the inner and outer borders of the iris and doing a one dimensional segmentation of the iris in a polar domain.

38. The method of claim 37, further comprising converting the features of the iris into an iris numeric code.

39. The method of claim 38, further comprising comparing the iris numeric code with previously generated iris numeric codes for verification and identification of the iris numeric code or for entry of the iris numeric code into a database.

40. The method of claim 37, wherein the inner and outer borders of the iris are approximated with ellipses.

41. The method of claim 40, wherein least squares modeling is used to estimate elliptic parameters and orientation.

42. The method of claim 40, wherein the ellipses are not necessarily concentric.

43. The method of claim 37, wherein the outer border of the iris is approximated with irregular edges due to eyelids and eyelashes.