基于DSP的疲劳驾驶实时监测系统研究

较为理想的非接触式检测眼睛疲劳程度的方法是通过摄像技术获取眼部的动态图像,利用累积差分帧和Hough变换等实时图像处理技术来检测、跟踪眼睛,分析眼睛的状态和提取眼睛的特征参数,在一定的时间内连续统计眼睛的闭合时间,计算出PERCLOS值来判断疲劳程度并采取相应的措施。把眼睛检测及跟踪算法移植到专用DSP上,最终开发出疲劳驾驶的嵌入式实时监测系统。实验结果表明,本系统的算法简单、快速、鲁棒性强。

Research on Fatigue Driving Real-time Monitoring System Based on DSP

Dynamic picture of eye obtained with video technology is a comparatively ideal non-contact measuring eyes tired degree methods,and utilize ADF and Hough transform technology of the real-time images to detect,track and analyze the state of eyes and pick-up characteristic parameter of eyes,count closing time of eyes in succession within regular hour,judge the tired degree and take the corresponding measure in value of calculating out PERCLOS.Detection and track algorithm of eyes are transplanted the special-purpose DSP,develop the embedded real-time monitoring system of fatigue driving finally.Experiments show that the algorithm is simple,fast and robust.