Method of detecting vehicle-operator state

• US 7,423,540 B2
• Filed: 12/23/2005
• Issued: 09/09/2008
• Est. Priority Date: 12/23/2005
• Status: Active Grant

• Current Assignee: Aptiv Technologies AG
• Original Assignee: Delphi Technologies, Inc. (BorgWarner Incorporated)
• Inventors: Kisacanin, Branislav
• Primary Examiner(s): Bugg; George A
• Assistant Examiner(s): Tang; Son M
• Application Number:
  US11/317,431
• Publication Number:
  US 20070159344A1
• Time in Patent Office: 991 Days
• Field of Search: 340/576, 340/937
• US Class Current: 340/576
• CPC Class Codes: G06T 2207/30201 Face
• CPC Class Codes: G06T 7/73 using feature-based methods
• CPC Class Codes: G06V 40/10 Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
• CPC Class Codes: G08B 21/06 indicating a condition of sleep, e.g. anti-dozing alarms

Abstract

A method of detecting the state of an operator of a vehicle utilizes a low-cost operator state detection system having no more than one camera located preferably in the vehicle and directed toward a driver. A processor of the detection system processes preferably three points of the facial feature of the driver to calculate head pose and thus determine driver state (i.e. distracted, drowsy, etc.). The head pose is generally a three dimensional vector that includes the two angular components of yaw and pitch, but preferably not roll. Preferably, an output signal of the processor is sent to a counter measure system to alert the driver and/or accentuate vehicle safety response.

Claims

1. A method for determining vehicle operator state for alerting the operator and comprising the steps of:

generating a non-stereo video image of said operator from a single camera, said image comprising a right eye, a left eye and a nose;

acquiring a first point, a second point and a third point of said image by a video processor, wherein the first point corresponds to the right eye, the second point corresponds to the left eye and the third point corresponds to the nose;

establishing a first line through the first and second points;

determining a cyclop'"'"'s point centered between the first and second points and lying on the first line;

establishing a second line drawn through the cyclop'"'"'s point and disposed perpendicular to the first line;

calculating a shortest squared second distance between the third point and the second line;

utilizing the second distance to calculate a horizontal component of a head pose by the video processor;

calculating a shortest squared first distance between the third point and the first line; and

utilizing the first distance to calculate a vertical component of the head pose.

2. The method for determining vehicle operator state set forth in claim 1 further comprising the steps of:

determining a nominal vertical component through a statistical learning process; and

utilizing the nominal vertical component to further calculate the vertical component of the head pose.

3. The method for determining vehicle operator state set forth in claim 2 wherein the nominal vertical component is generally the most common vertical component of the head pose.

4. The method for determining vehicle operator state set forth in claim 3 wherein the statistical learning process omits about the first minute of driving time to eliminate any excessive motion of the operator and the driver is assumed not to be drowsy or distracted.