Dynamic eye tracking system

• US 7,742,621 B2
• Filed: 06/13/2006
• Issued: 06/22/2010
• Est. Priority Date: 06/13/2006
• Status: Active Grant

• Current Assignee: Aptiv Technologies AG (Aptiv PLC)
• Original Assignee: Delphi Technologies, Inc. (BorgWarner Incorporated)
• Inventors: Hammoud, Riad I., Malawey, Phillip V
• Primary Examiner(s): Ahmed, Samir A
• Assistant Examiner(s): HU, FRED H.
• Application Number:
  US11/452,116
• Publication Number:
  US 20070286457A1
• Time in Patent Office: 1,470 Days
• Field of Search: 382/103, 382/117, 382/118
• US Class Current: 382/103
• CPC Class Codes: A61B 3/113 for determining or recording eye movement
• CPC Class Codes: A61B 5/163 by tracking eye movement, gaze, or pupil change
• CPC Class Codes: A61B 5/18 for vehicle drivers or machine operators
• CPC Class Codes: G06F 3/013 Eye tracking input arrangements G06F3/015 takes precedence
• CPC Class Codes: G06V 20/597 Recognising the driver's state or behaviour, e.g. attention or drowsiness
• CPC Class Codes: G06V 40/19 Sensors therefor
• CPC Class Codes: G08B 21/06 indicating a condition of sleep, e.g. anti-dozing alarms

Abstract

A dynamic eye tracking system to track movement of an eye including the steps of detecting if a neighbor exists about the eye over a predetermined amount of time and forming a boundary for the neighbor. Eye candidates are tracked and a determination is made if the eye candidates are beyond the boundary defined by the detected neighbor.

Claims

1. A dynamic eye tracking method to track movement of an eye in video images from a camera, said method comprising:

providing a camera;

using a computerized eye state monitor having a computer readable medium;

detecting if a neighbor exists about the eye using initial coordinates of the eye for over a predetermined amount of time within a predefined area;

forming a boundary for said neighbor within said predefined area;

tracking potential eye candidates; and

determining if said eye candidates are beyond the boundary and eliminating the eye candidates outside the boundary defined by the detected neighbors within said predefined area.

2. A dynamic eye tracking method as defined in claim 1 further comprising:

determining if there is a persistent neighbor above said eye in successive video images; and

determining if the eye image is merged with an image of said neighbor around said eye, and if so offsetting a previously known boundary of the neighbor by the change in the eye position to update said boundary for said merged neighbor image.

3. A dynamic eye tracking method as defined in claim 2 further comprising:

comparing a local inverse receptive hole extraction for candidates not past said boundary and a local iris candidate not past said boundary to provide an eye candidate if there is no persistent neighbor above said eye or said eye is not merged with said neighbor above said eye.

4. A dynamic eye tracking method as defined in claim 3 further comprising:

using a single template correlation by comparing with pre-loaded acceptable images of eyes to determine if there is an eye candidate if said local inverse receptive hole extraction and said local iris extraction provides no acceptable eye candidate, and to determine if said eye provided is beyond the boundary determined by said neighbor.

5. A dynamic eye tracking method as defined in claim 4 further comprising:

comparing previously saved templates from above, below, to the left, and to the right of the eye and find a predetermined minimum correlation and execute an appearance check at the intersection of a vertical line between above and below templates and a horizontal line between left and right eye templates.

6. A dynamic eye tracking method as defined in claim 1 further comprising:

executing in sequence any one of a series of eye candidate subroutines to determine if an acceptable eye candidate exists and for each subroutine then determine if said eye candidate is beyond or within a boundary defined by said neighbors within a predefined area.

7. A dynamic eye tracking method to track movement of an eye in video images from a camera, the method comprising:

providing a camera;

using a computerized eye state monitor having a computer readable medium;

executing in sequence any one of a series of eye candidate subroutines to determine if an acceptable eye candidate exists; and

for each subroutine then determining if said eye candidate is beyond or within a boundary defined by said neighbors within a predefined area.