The Research About Face Research About Face Fatigue Detection Technology Based On Adaboost Algorithm

Traffic accidents killed tens of thousands of people every year in our country, and cause huge economic losses.On account of the fatigue driving is one of the main cause of traffic accidents,so the technology of fatigue driving detection has become a research hotspot at home and abroad in recent years.To work out the system which can detect fatigue effectively and remind driver promptly is of great significance to protect the personal safety and reduce the country’s economic loss.Based on the study of the fatigue detection technology at home and abroad,in view of the characteristics of facial fatigue, this paper put forward the method of using PERCLOS, blinking rate and head position to detect fatigue,the system timely warn when detect fatigue, the main content of this paper is as follows:First, in terms of face detection, this paper combine the Adaboost face detection with skin color validation method to improve the accuracy of the system. Through the analysis of the color space clustering, this article choose the YCbCr color space to test the the candidate face area which is detected by Adaboost method.Second, after the analysis of advantages and disadvantages of the various human eye detection method, this article combine face edge projection with Adaboost eye detection to locate eye, using edge projection to locate the human eye roughly, then apply Adaboost detector in the precise location of human eyes, and on this basis, chooses the kalman filtering to track the human eye in real time.Finally, adaptively transform the eye area into binary chart,the state of eye is judged by calculating the area of the eye,then make use of PERCLOS,blink rate and nod rate based on eye position to detect fatigue.