Design And Research Of Fatigue Driving Monitoring System Based On ARM And Linux

With the booming development of the economy, the number of cars has been increased rapidly. At the same time, the road traffic accidents are increasing day by day. According to statistics, up to92percent of accidents involving casualties are caused by drivers’problems, and most accidents caused by fatigue driving always are serious big accidents. Only in2011,27major accidents took place nationwide and killed451people, one of which caused more than10deaths. Traffic accidents will be sharply reduced and lives saved if fatigue driving is effectively avoided. So it has very important theory significance and practical value, that we reaches the drivers’behavior characteristics through the Meidicine, Psychology and other related knowledge, and designs the monitoring system of fatigue driving using the computer, signal processing, intelligent control and other science and technology.This issue is further researches and improvements made on the basis of the project of my tutor and seniors. It is intelligent Monitoring Device of Motor Vehicle Driving State> completed during2007to2008(not perfect), which is a projiect of promotion of achievement from the Shanxi Science and Technology Department. The previous seniors have researched the Monitoring System of Fatigue Driving separately based on one of three different aspects:the drivers’ physiological signal, the drivers’external behavior and the parameters of vehicle. There are some limitations due to a single signal acquisition. Last senior build the embedded hardware platform of monitoring system of fatigue driving. He has designed the hard circuit with the core processor of S3C2440A, and proved the feasibility.On the basis, improvements have been made in this paper. The main research works of this paper are as follows.(1) A brief description of the hardware platform of ARM9, with a S3C2440A of SAMSUNG as its core processor. Analysis the interface circuits of memory module, JTAG module, serial port module, USB and CAN-bus.(2)Select the Embedded System-Linux as operating system. According to the characteristics of this system, build the cross compile environment, design and implement the transplantation of Bootloader, transplantate the Linux kernel, develop the Device driver and built the root file system.(3)Study on the detection method of drivers’faces. Establish skin color model using the good clustering of face color in the YCbCr space. Then determine the color region through similarity calculation. Finally locat the driver’s face through the face features.(4) Study on the detection method of drivers’eyes. Extract the eyes’ contour using the improved Snake algorithm and locat the eyes. Extracte and analysis the informations related with fatigue:PERCLOS value, eyes closed time period and blink frequency, to obtaine the fatigue driving indicators.(5)Study on the detection method of drivers’pulse. Through the improved Welch method make power spectrum estimation for the pulse signal. Extracte and analysis the informations related with fatigue:power spectrum peak value and peak frequency, to obtaine the fatigue driving indicators.(6)Study on fatigue driving monitoring methods based on Multi-information fusion. This paper applies Rough Set theory into the system. Firstly, discrete the information collected. Secondly, obtain the minimal decision algorithms through the attribute reduction. Finally, judge the degree of driver’s fatigue using the decision algorithms.On the basis of our researches, we declare a projiect,<Research and Design of Intelligent Monitoring System of Fatigue Driving>, to Shanxi Adminstration of Work Safety this year, which is to be approved.