Driving Fatigue Monitoring System Based On Jz4755 Implementation

With the rapid development of transportation technologies, the occurrence of traffic accidents are also constantly increased, and which posed a serious threat to people’s lives, property and lives. The proportion of traffic accidents caused by driver fatigue is very large, therefore, preventing fatigue driving can effectively reduce traffic accidents. However, there has no mature products about the system of driver fatigue monitoring on the market currently, the main reasons:First, some algorithms are too immature to be exploited;Second, the applicability and robustness of the system is a little weak that they can not be dealt with the complex driving environment. A large number of studies have shown that monitor the state of fatigue of the human eye has an absolute advantage in driving fatigue monitoring methods, papers based on this approach, a driver fatigue monitoring system to achieve real-time monitoring of the status of the human eye based on the JZ4755is posed.Based on domestic and foreign excellent fatigue monitoring plans. Papers proposed an real-time driver fatigue monitoring system which is based on the ARM chip of JZ4755. Details are as follows:(1) Paper summarized the research’s background, significance, and research status, then analyzed the trends. It is introduced the interrelated knowledge of the eye fatigue monitoring system, then introduced digital image denoising, enhancement, feature extraction, image segmentation and other related algorithms. And then, Described the remodel plan of the CCD camera which can get the images with bright pupil effect which is produced by switching the near-infrared wavelengt of850nm and940nm’s LED. At last introduced the principle of PERCLOS parameters and the specific application in this system.(2) It summarized some typical algorithm of face recognition at home and abroad, and introduced the principle of the AdaBoost face detection algorithm which is based on Haar features, and introduced the training steps and the plan which is used in the system. Then,it summarized the typical human eye localization algorithm, and introduced the principles and steps of the kalman filter tracking algorithm.(3) Paper discussed the algorithm of the system:The images with bright pupil effect are collected by the the CCD camera which controlled the LED lights with different infrared wavelength alternately switching by MCU8051. Then, the eye area is coarse located by gray integral projection after median filter. Then, the original eye area image is binarization processed by the method of Otsu, then, picked up the eye contour by the method of Sobel, and then decided the condition of driver fatigue by calculating parameters of PERCLOS. After precisely targeted the eyes, tracking eyes’area by kalman filter tracking algorithm.(4) Selecting and improving the algorithm by the experiments with matlab, and the experimental results are given then. The hardware design of the system is introducesd, and discusses the architecture and function of the chip of JZ4755with the400MHz of the CPU, and introduced the development methods and procedures in the system based on the chip through linux OS. Introduced the main functions of the drive safety system:driver fatigue monitoring, prevention of blindness area, driving records and GPS navigation and so on.