Real time driver fatigue warning system based on multi-source information fusion

Abstract

A real time driver fatigue warning system based on multi-source information fusion. Through the real-time acquisition of the driver's facial image from the infrared camera, median filtering of the face image and histogram equalization preprocessing are carried out. After the localization of the initial frame of human eye, the initial frame of the mouth is positioned, and then the real-time tracking is carried out. The OSTU algorithm was used to analyze the state of the eyes and mouth, and the state of the eyes and mouth was analyzed. Finally, the alarm system was activated by wireless transmission equipment when it is necessary according to the different levels of fatigue. The invention has the advantages of small overall operation amount, fast transmission rate, high detection precision and fast response speed, and has the advantages of more practicality and wider application range.

Description

technical field [0001] The invention relates to the technical field of driver fatigue early warning, in particular to a real-time driver fatigue early warning system based on multi-source information fusion. Background technique [0002] According to the Yearbook of Traffic Accidents: Fatigue driving often leads to serious traffic accidents, and 17% of fatal traffic accidents are caused by fatigue driving. However, drivers generally do not realize that they are driving fatigued before the accident actually occurs. Fast and accurate human eye and mouth positioning and tracking algorithms can improve the detection speed of fatigue driving status, which is of great significance to reduce the occurrence of traffic accidents. The fatigue detection system analyzes the fatigue state of the driver in real time, and sends corresponding early warning information when necessary according to the different levels of quantified fatigue. It can effectively avoid the occurrence of traffic ...

AI Technical Summary

Problems solved by technology

[0003] The combination of traditional Kalman filter and mean shift algorithm can track the target, but it often fails to track when the target is fast moving and severely occluded. At the same time, the time cost of the algorithm cannot be ignored, and it cannot meet the real-time requirements.

[0004] In the prior art, grayscale projection is sensitive to face wearing glasses and posture changes, and the template matching algorithm has high precision, but the calculation amount is too large, the procedure is cumbersome, and it is very inconvenient

[0005] The traditional template matching is to compare the pixel gray value or color value of the same position between the whole image and the target image. This method has the problems of high computational complexity, sensitivity to target rotation, and many interference factors.

Images