Real-Time Detecting System Of The Driver's Fatigue

Driver inattention, particularly due to fatigue, is widely recognized as one of the most causes in accidents. Studies have shown that drivers often gape when tired and the frequency of blinking increases. So if we can detect the facial changes of the drivers and provide prompt warning when necessary, then we can avoid the accidents caused by the fatigue of drivers.The appearance of the high-speed processor in recent years has made the real-time detection of the video possible. This provides us with hardware support. The system based on the TMS320C6416 DSP processor with a timing frequency of 600MHZ produced by the TI Corporation. The infrared CCD camera collects the image signal and DSP processes it, then the result is shown on the LCD.This system applies DSP/BIOS to build two tasks and two software interruptions. One task initializes the board, the other analyzes and processes the collected video signal. If it finds out that the driver is in a state of fatigue, it will awaken the software interruption immediately to give out phonetics warning. The prior of eyes warning is prior than the mouth warning.With the improving function of DSP chip, the research emphases of signal processing system has again transferred from hardware to software arithmetic, especially in the area of real-time image processing. This system is to detect whether the drivers are in the state of fatigue. It should be built in a certain environment, namely, in the cab of the drivers. Then we can get an image mainly of the drivers' head through adjusting the installing position and the focus of the vidicon.We first need to make an exact orientation of the driver's face. The object in the world that the color of the face and the outline of the head are similar to human is more or less rarely exists. So the integrated method of differential frames arithmetic and complexion detection can make an instant and exact orientation of the driver's face. On the basis of the confirming face, we use many methods to detecting the driver's eyes and mouth. Finally, we make an analysis of whether the driver is in the state of fatigue. If he is, there will be an incurrence of phonetic warning. This paper takes full advantage of each method, when comes to the defects of them, we caneither apply another method to make a complement, or revise the terms affecting the arithmetic. Make sure the detection is high in accuracy and strong in real-time attribute.At present, the traffic accident due to the fatigue driving is increasing year by year. So the method of the image examination to driver's driving state is better to solve the problem. This hardware platform of the system can well meet the real-time processing demand for the image and audio frequency, the configuration of the system is felicitousness and the algorithm is high in veracity and strong in real time attribute. The system this article puts can accurately examine the driver's driving state and when he is tired sending out the phonetic warning.