According to the U.S. Joint Investigation Bureau, 20%~30% of Giti accidents are caused by fatigue driving, and the research on driver fatigue detection technology is of great significance for the prevention of Giti accidents. In the recent decade, fatigue driving detection has gradually gained people's attention, and the current research methods of driver fatigue driving detection can be divided into two major categories: one is to indirectly judge whether the driver is fatigued according to the behavioural performance of the vehicle. In this type of technology, through the sensor to obtain the vehicle in the driving process of a variety of parameters, according to the vehicle whether the dog road marking line, speed whether the speed, whether the distance between the vehicles is too close, etc., to determine whether the driver has fatigue. The second from the driver's own characteristics, through some kind of equipment to obtain the driver's physiological parameters characteristics or visual characteristics, the use of drivers in the normal state and the fatigue state of the characteristics of different patterns, the use of the corresponding pattern recognition technology classification for the discrimination, so as to detect whether there is fatigue.

Fatigue detection technology is using intelligent video analysis technology, including the driver's face detection and human eye detection and analysis, in order to determine the fatigue level of the detected person, and according to the set rules for fatigue alarms and warnings.

Fatigue detection system, mainly consists of fatigue detection special camera and fatigue detection terminal host, through the installation of the camera on the driving dashboard to directly shoot the driver's face, and accurately locate the driver's eyes, the system will automatically determine and analyse the closing state of the human eye and the frequency of the closure, the judgement of the driver's face towards the direction of the vehicle, and at the same time, combined with the current vehicle speed, and according to the things set up to detect the Logic rules, based on a variety of parameters for the analysis to determine the driver's fatigue state, for example: when at a certain speed moment, the length of time the eyes are closed to reach the alarm value, or when the driver's face more than a certain period of time without facing forward (such as bowing the head for 3 seconds) the situation occurs, the system will automatically send out an alarm sound, prompting drivers to misoperation and fatigue driving, when the situation is seriously, the system will also be through the network will be alarmed! When the situation is serious, the system will also notify the backstage monitoring centre through the network to supervise the driver by manual intervention.

Humanised design:　　

Considering that most drivers have the habit of driving with sunglasses, and fatigue detection of special cameras using special wavelengths of infrared light complementary lens, can effectively penetrate the driver's sunglasses, and will not be unable to use the fatigue detection system due to the driver wearing sunglasses.

Detection rules are set:　　

Fatigue detection warning, need to refer to the vehicle's current driving speed, the driver's eyes closed degree, the driver's face orientation and other parameters, and can be customised from the actual need to set the alarm limit, higher flexibility, can be based on different vehicles, different driving sections, etc. to set the parameters, convenient for users to use.