基于图像处理的驾驶员疲劳检测系统

Image Processing-based Warning System for Fatigue Driver

为了防止驾驶员疲劳驾驶,以驾驶员的眼睛作为研究对象,采用图像处理技术设计仿真了一个基于Matlab的驾驶员疲劳检测系统。该系统采用自带光源的摄像头采集序列图像,对图像中的眼睛点进行定位跟踪,然后计算出眼睛面积,最后对驾驶员是否进行疲劳驾驶做出判断并进行预警。仿真测试结果表明,该系统对驾驶员浅度疲劳的识别率为81.5%,对中度疲劳和深度疲劳的识别率为100%。

In order to prevent fatigue driving,through many studies of the driver' eyes,a Matlab-based fatigue driving detection system is designed and simulated based on image processing technology.The system samples sequences of images with an camera and its built-in light source,and then locates the eye-point in the image and tracks it,calculates the area of the eyes,makes judgment on whether the driver is driving with fatigue or not and warns him according to the judgment.The simulation result shows that the recognition ratio of shallow fatigued drivers is 81.5 %,and the recognition ratio of moderate fatigued drivers and deep fatigued drivers is 100 %.