Driver fatigue detecting method and system based on vision

The invention relates to a driver fatigue detecting method and a system based on vision, belonging to the technical field of computers. The method comprises an image acquisition part, a human eye detection part, a fatigue judgment part and a fatigue alarming part. In the image acquisition part, a camera is used to acquire the facial image of a driver and the image is transmitted to a mobile phone end by a local area network constructed by a wireless router; in the human eye detection part, adaboost algorithm is used to carry out human face detection and human eye detection and the eye state can be judged by human eye area binarization; in the fatigue judgment part, the fatigue degree can be calculated by combining frequency of wink with the P80 measurement method of PERCLOS according to the open and close states of the human eyes; and in the fatigue alarming part, and when the fatigue state is judged, a mobile phone alarms by ringing and simultaneously triggers an alarm to alarm. By the method and the system disclosed by the invention, alarm signals can be sent in time when a driver is in fatigue driving, and the safety of life and property of drivers is guaranteed. The method and the system can be as applied to general driving safety accessories and traffic supervision and the like, and also can be used for fatigue detection of pilots and the crew. The method and the system disclosed by the invention has high reliability, low cost and good applicability.