摘要： 利用红外敏感的摄像机获得司机脸部图像，通过可变性模板匹配的方法对眼睛进行定位，然后利用卡尔曼滤波的方法跟踪眼睛，得到司机的眨眼持续时间参数，以此为依据判断司机是否疲劳。主要研究了其中的图像处理方法——特征提取算法和眼睛定位、跟踪算法。实验结果证明，眨眼持续时间判断是否疲劳的有效指标。

关键词: 眨眼持续时间, 可变性模板, 卡尔曼滤波

Abstract: This paper uses infrared camera to obtain the image of the driver’s face and locate their eyes with a deformable template in the image. It uses Kalman filter to track the eyes and obtain the parameters of the blink duration which are used to judge the fatigue of the driver, and researches the image processing methods in the algorithm including feature extraction, eyes location and tracking. Experimental results show that blink duration is an effective parameter to judge fatigue.

Key words: blink duration, deformable template, Kalman filter

中图分类号: 

• TP391.41