System for monitoring eyes for detecting sleep behavior

• US 5,570,698 A
• Filed: 06/02/1995
• Issued: 11/05/1996
• Est. Priority Date: 06/02/1995
• Status: Expired due to Term

• Current Assignee: Siemens Medical Solutions USA Incorporated (Siemens AG)
• Original Assignee: Siemens Corporate Research Incorporated (Siemens AG)
• Inventors: Fang, Ming, Liang, Cheng-Chung, Singh, Ajit
• Primary Examiner(s): Sykes, Angela D.
• Assistant Examiner(s): CARTER, RYAN CHRISTOPHER
• Application Number:
  US08/459,148
• Time in Patent Office: 522 Days
• Field of Search: 128/745, 351/206, 351/209, 351/211, 351/221, 340/573, 340/575
• US Class Current: 600/558
• 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: G06V 20/597 Recognising the driver's state or behaviour, e.g. attention or drowsiness
• CPC Class Codes: G06V 40/193 Preprocessing; Feature extraction

Abstract

A system for monitoring eyes of an individual includes a human interface for obtaining images of the individual and for providing feedback to the individual and a processor for analyzing the images. The processor transforms an image sequence into a one dimensional signal by extracting relevant features from the images. Analysis of the signal generated then occurs to detect sleepiness. Transformation of the image sequence includes eye localization, eye tracking and eye motion signal generation. The system takes advantage of the relatively high horizontal-contrast density of the eye region to determine eye positions in a greyscale image of a human face.

Claims

1. A system for monitoring eyes for detecting sleep behavior comprising:

human interface means for non-invasive monitoring of said eyes of an individual, for providing scanned images and for providing feedback to said individual; and

,

processor means connected to said human interface means for receiving said scanned images from said human interface means, for processing said scanned images and for providing processed data back to said human interface means for providing said feedback to said individual, wherein said human interface means comprises;

imaging means connected to said processor means for monitoring said eyes of said individual; and

,

warning system means connected to said processor means for receiving said processed data from said processor means and for providing a warning to said individual, and wherein said processor means comprises;

eye localization means connected to said human interface means;

eye tracking means connected to said eye localization means;

eye motion signal generation means connected to said eye tracking means; and

,

analysis means connected between said eye motion signal generation means and said human interface means.

2. A system for monitoring eyes for detecting sleep behavior as claimed in claim 1 wherein said eye localization means comprises:

horizontal-contrast computation filter means;

horizontal-contrast density determination filter means connected to said horizontal-contrast computation filter means;

facial geometry reasoning means connected to said horizontal-contrast density determination filter means; and

,

eye position determination means connected to said facial geometry reasoning means.

3. A system for monitoring eyes for detecting sleep behavior as claimed in claim 2 wherein said eye tracking means comprises:

search means for correlating an image-block with an image in a current frame at varying displacements from an original location.

4. A system for monitoring eyes for detecting sleep behavior as claimed in claim 3 wherein said eye motion signal generation means comprises:

measurement means for measuring correlation between a new tracked image block and a very first initial block where an eye is open.

5. A system for monitoring eyes for detecting sleep behavior as claimed in claim 4 wherein said analysis means evaluates a tracking T curve and a frame difference D curve.

6. A system for monitoring eyes for detecting sleep behavior as claimed in claim 5 wherein said processor means further comprises:

greyscale correlator means connected to said human interface means.

7. A system for monitoring eyes for detecting sleep behavior as claimed in claim 6 wherein said processor means further comprises:

video display chip means connected to said human interface means.

8. A system for monitoring eyes for detecting sleep behavior as claimed in claim 7 wherein said human interface means further comprises:

controller means connected to said imaging means for allowing said individual to adjust viewing angle and direction of said imaging means.

9. A system for monitoring eyes for detecting sleep behavior as claimed in claim 8 wherein said human interface means further comprises:

track/lost-track indicator means for informing said individual that said imaging means needs adjustment.

10. A system for monitoring eyes for detecting sleep behavior as claimed in claim 9 wherein said imaging means comprises:

a CCD camera.

11. A method of monitoring eyes for detecting sleep behavior comprising the steps of:

noninvasively monitoring said eyes of an individual;

providing scanned images;

processing said scanned images; and

,

providing feedback to said individual, wherein the step of processing said scanned images comprises the steps of;

localizing said eyes;

tracking said eyes;

generating eye motion signals; and

,

analyzing said eye motion signals, and wherein the step of localizing said eyes comprises the steps of;

horizontal-contrast computation filtering of a greyscale image of a face for providing a binary mask image;

horizontal-contrast density determination filtering of said binary mask image for providing a greyscale mask image;

performing facial geometry reasoning on said greyscale mask image for providing approximate positions of two eyes; and

,

performing eye position determination of said greyscale image of a face and said approximate positions of two eyes for providing positions of two eyes.

12. A system for monitoring eyes for detecting sleep behavior comprising:

human interface means; and

,

processor means connected to said human interface means wherein said processor means comprises;

greyscale correlator means connected to said human interface means;

eye localization means connected to said greyscale correlator means;

eye tracking means connected to said eye localization means;

eye motion signal generation means connected to said eye tracking means; and

,

analysis means connected between said eye motion signal generation means and said human interface means.

13. A system for monitoring eyes for detecting sleep behavior as claimed in claim 12 wherein said human interface means comprises:

imaging means connected to said greyscale correlator means for monitoring said eyes of an individual; and

,

warning system means connected to said analysis means for receiving analyzed data from said analysis means and for providing a warning to said individual.

14. A system for monitoring eyes for detecting sleep behavior as claimed in claim 13 wherein said human interface means further comprises:

controller means connected to said imaging means for allowing said individual to adjust viewing angle and direction of said imaging means; and

,

track/lost-track indicator means for informing said individual that said imaging means needs adjustment.

15. A system for monitoring eyes for detecting sleep behavior as claimed in claim 15 wherein said processor means further comprises:

video display chip means connected to said human interface means.

16. A system for monitoring eyes for detecting sleep behavior as claimed in claim 19 wherein said eye localization means comprises:

horizontal-contrast computation filter means;

horizontal-contrast density determination filter means connected to said horizontal-contrast computation filter means;

facial geometry reasoning means connected to said horizontal-contrast density determination filter means; and

,

eye position determination means connected to said facial geometry reasoning means.