The Study On The System For Real-time Detecting Fatigue Of Drive Based On DSP

It is statistic that over 50% in traffic accident caused in thruway are result from drowsiness and absence, even nap because of longtime driving. Drowsy driver detection methods can form the basis of a system to potentially reduce accidents related to drowsy driving. Drowsy driver detection methods mainly centralize on the three aspects: based on character of driver;based on physiological parameter of drive;based on character of vehicle. However, the last two of methods are too restricted to vehicle or experience and performance of drive, or are invasive for the driver. Office of Motor Carrier Researcher and Standards investigate the potential of an eye tracking system for detecting reduced driver alertness, and to determine the impact of preventative napping on driver alertness and performance. Partial eye closures during fixations, measured as the ratio of vertical to horizontal pupil diameter (V/H), used as a measure of drowsiness-detection effectively. The results show this parameter is particularly impressive as an indicator of degraded alertness at least 2 to 3 minutes before an accident and likely as much as 10 to 12 minutes before an accident.According to it, I design a system for real-time detecting fatigue of driver. Using an infrared CCD camera to track the driver, the system based on DSPTMS3206416 can execute arithmetic in real time. Also it can perform effectively on daytime or nighttime.How to locate the eyes is a linchpin to the system. Wavelet transform and neural networks for pattern recognition as promising approaches, these routines had severe difficulty actually detecting faces, most likely due to insufficient or improper training, and too complicated to carry out. Equal-area map projection, template matching, Hough translation are easily used to detect fatigue of drivers. First locate the vertical position of chin, mouth, nose, and eyes with gray-level projection, then localizes the horizontal position of them in associated vertical positions. The eyes can be approx located. Then, it can be exactly located based on approximate position at second time.