在司机疲劳监控中的人眼状态检测方法研究

Research of Eye State Identifying in Fatigue Detecting and Controlling System for Driver

主要研究了在确定了司机眉眼区域的基础上进行人眼定位和状态检测的方法,首先将获取的眉眼区域图像分为4部分,然后分别计算复杂度并选取其中复杂度最大的部分,对其进行积分投影和方差投影相结合的混合投影,最后根据几何特征来判断人眼的状态。

A method of locating the eye area and detecting eye state on the basis of that the image of the eyebrow and eye area of drivers is introduced.Firstly,the image of eyebrow and eye area is separated to four parts,and the most complex part is chosen after calculating the complexity of each part.Then,the combined projection algorithm which combined with integral projection and differential projection is used to it.Finally,the state of eye can be identified according to the geometry character.