Method and apparatus for detection of drowsiness and quantitative control of biological processes

• US 20080192983A1
• Filed: 12/20/2007
• Published: 08/14/2008
• Est. Priority Date: 10/28/2002
• Status: Active Grant

• Current Assignee: Morris Steffin
• Original Assignee: Morris Steffin
• Inventors: Steffin, Morris
• Granted Patent: US 7,680,302 B2
• Time in Patent Office: Days
• US Class Current: 382/104
• CPC Class Codes: B60K 28/06 responsive to incapacity of driver
• CPC Class Codes: G06T 7/0012 Biomedical image inspection
• CPC Class Codes: G06V 40/168 Feature extraction; Face representation
• CPC Class Codes: G06V 40/171 Local features and components; Facial parts eye characteristics G06V40/18; Occluding parts, e.g. glasses; Geometrical relationships
• CPC Class Codes: G08B 21/06 indicating a condition of sleep, e.g. anti-dozing alarms

Abstract

The present invention is directed to a real-time automated video cognizer that functions as a facial video processor for detecting drowsiness in operators of motorized vehicles, including the use of a video cognizer to provide pattern recognition and control signal generation during monitoring of macroscopic or microscopic biological processes. More specifically, the present invention accepts input from a video monitoring system that continuously captures the operator'"'"'s facial images; employing three sequential means of processing the digitized video information to extract the position and configuration of drowsy-relevant facial features and numerically processes this information to yield a quantitative estimate of drowsiness probability in each epoch of monitoring. The means of the present invention are noninvasive, do not restrict driver movement or performance, provide increased measurement reliability for actual driver behavior, and include the capability of generating or triggering suitable alarms when drowsiness occurs.

Claims

1. A method for monitoring a biological process, said method comprising:

receiving image data of an object having one or more features, said image data corresponding to frames comprising a plurality of pixels;

dividing each frame into one or more regions based on the one or more features and each region into one or more subregions;

filtering pixels in each subregion according to a pixel intensity range to provide a filtered output of pixels;

defining for each frame a first area within each subregion, said first area defined by one or more predetermined shape equations according to a shape of a feature to be monitored in the subregion; and

evaluating for each frame the number of filtered output of pixels in the first area for each subregion to transform the image data to a scalar feature signal for each subregion.

2. The method of claim 1 further comprising:

defining for each frame a second area within each subregion, said second area adjacent to said first area; and

wherein

said step of evaluating comprises determining for each frame the difference between the number of filtered output pixels in the first area and the number of filtered output pixels in the second area and outputting a value of the difference to produce the scalar feature signal for each subregion.

3. The method of claim 1 further comprising:

detecting one or more changes in a feature in a first area of a subregion; and

modifying parameters of said shape equations defining said first area to correspond to the detected changes.

4. The method of claim 2 further comprising:

transforming each scalar feature signal to produce a behavior indicating output signal for each subregion.

5. The method of claim 4 further comprising:

combining the behavior indicating output signals for each subregion to obtain a composite behavior indicating output signal to monitor the biological process.

6. The method of claim 5 wherein said step of combining comprises linearly combining the behavior indicating output signals.

7. The method of claim 5 further comprising:

correlating the composite behavior output signal with one or more independent measures of the biological process so as to increase the accuracy with which the composite behavior output signal monitors the biological process.

8. The method of claim 5 further comprising:

modifying the composite behavior indicating output signal pursuant to an algorithm under program control to produce a composite measure of the biological process; and

determining whether the composite measure is below a threshold.

9. The method of claim 1 further comprising acquiring image data via computerized microscopy;

and

wherein the biological process is microscopic at the tissue, cellular or subcellular level.

10. The method of claim 1 wherein the biological process is macroscopic.

11. The method of claim 10 wherein the macroscopic biological process is drowsiness.

12. The method of claim 1 wherein the one or more regions comprises an eye region, a mouth region and a facial boundary region.

13. The method of claim 1 wherein said step of filtering comprises:

determining whether a video intensity level of each pixel is within the pixel intensity range; and

setting the video intensity level to a predetermined value if the video intensity level is within the range and to another predetermined value if the video intensity level is outside the range to provide the filtered output.

14. An apparatus for monitoring a biological process, said apparatus comprising a processor programmed to perform a method, said method comprising:

receiving image data of an object having one or more features, said image data corresponding to frames comprising a plurality of pixels;

dividing each frame into one or more regions based on the one or more features and each region into one or more subregions;

filtering pixels in each subregion according to a pixel intensity range to provide a filtered output of pixels;

defining for each frame a first area within each subregion, said first area defined by one or more predetermined shape equations according to a shape of a feature to be monitored in the subregion; and

evaluating for each frame the number of filtered output of pixels in the first area for each subregion to transform the image data to a scalar feature signal for each subregion.

15. The apparatus of claim 14 wherein the processor is programmed to perform the method further comprising:

defining for each frame a second area within each subregion, said second area adjacent to said first area; and

wherein

said step of evaluating comprises determining for each frame the difference between the number of filtered output pixels in the first area and the number of filtered output pixels in the second area and outputting a value of the difference to produce the scalar feature signal for each subregion.

16. The apparatus of claim 14 wherein the processor is programmed to perform the method further comprising:

detecting one or more changes in a feature in a first area in a subregion; and

modifying parameters of said shape equations defining said first area to correspond to the detected changes.

17. The apparatus of claim 15 wherein the processor is programmed to perform the method further comprising:

transforming each scalar feature signal to produce a behavior indicating output signal for each subregion.

18. The apparatus of claim 17 wherein the processor is programmed to perform the method further comprising:

combining the behavior indicating output signals for each subregion to obtain a composite behavior indicating output signal to monitor the biological process.

19. The apparatus of claim 18 wherein the processor is programmed to perform the method wherein said step of combining comprises linearly combining the behavior indicating output signals.

20. The apparatus of claim 18 wherein the processor is programmed to perform the method further comprising:

correlating the composite behavior output signal with one or more independent measures of the biological process so as to increase the accuracy with which the composite behavior output signal monitors the biological process.