Method and apparatus for detecting fatigue driving

Abstract

The invention discloses a fatigue driving detection method, which comprises the following steps: acquiring a head image by a near-infrared 3D camera; two consecutive head images are tracked by LK optical flow method, and mouth feature points, head feature points and eye feature points are extracted from the tracked head images; searching an eye image according to the position coordinates of the eye feature points to extract the eye feature points of the eye image; according to the eye feature points, mouth feature points and head feature points of the eye image within the predetermined acquisition times, the eyelid movement features, mouth movement features and head movement features being determined one-to-one, and fed into the SVM classifier, so that the SVM classifier judges the fatiguedriving state and realizes fatigue driving detection. The invention also provides a fatigue driving detection device. By adopting the fatigue driving detection method and the fatigue driving detection device of the invention, the accuracy and the judgment speed of the fatigue driving state judgment can be effectively improved.

Description

technical field [0001] The invention relates to the technical field of motor vehicle fatigue driving detection, in particular to a fatigue driving detection method and device. Background technique [0002] Driver drowsiness and fatigue are well-recognized risk factors for road safety. According to the report of the World Health Organization, more than tens of thousands of people are injured or even lose their precious lives due to fatigue driving by themselves or others every year. Therefore, alerting drivers to fatigue driving can effectively reduce traffic accidents caused by fatigue driving, and the detection of fatigue driving status plays a vital role in the timeliness and effectiveness of fatigue driving alerts. [0003] At present, the methods of fatigue driving detection mainly include the following four types: [0004] (1) Fatigue driving detection based on physiological signals: Relevant studies have shown that the physiological indicators of the driver in the fa...

AI Technical Summary

Problems solved by technology

[0008] At present, the fatigue driving detection method based on the state of the driver's head and facial features mainly acquires images through standard cameras. Since the frame rate of standard cameras is 25 to 30 Hz, and the blink duration may be very short (minimum about 150ms), so The number of sampling times for the video signal within one blink duration is only 4 to 5, which is not enough to obtain accurate eyelid movement characteristics, resulting in low accuracy and slow judgment speed for judging the fatigue driving state

Images