基于卷积神经网络的人眼状态检测  

Eye state detection based on convolutional neural network

作　　者：黄洁媛 岑翼刚[1] 张琳娜[2] 阚世超 梁列全[3] HUANG Jieyuan;CEN Yigang;ZHANG Linna;KAN Shichao;LIANG Liequan(School of Computer&Information Technology,Beijing Jiaotong University,Beijing 100044,China;School of Mechanical Engineering,Guizhou University,Guiyang 550025,China;Electronic Business Research Institute^Guangdong University of Finance&Economics^Guangzhou 510320,China)

机构地区：[1]北京交通大学计算机与信息技术学院,北京100044 [2]贵州大学机械工程学院,贵阳550025 [3]广东财经大学电子商务研究院,广州510320

出　　处：《扬州大学学报（自然科学版）》2018年第3期25-29,34,共6页Journal of Yangzhou University：Natural Science Edition

基　　金：国家自然科学基金资助项目(61572067);广东省自然科学基金资助项目(2016A030313708);北京市自然科学基金资助项目(4162050);广州市科技计划资助项目(201804010271)

摘　　要：提出了一种基于卷积神经网络(CNN)的人眼检测及眼睛睁闭状态分类的方法.首先,训练1个用于检测人眼中心点的卷积神经网络,当输入人脸图像时,网络能快速检测到双眼中心点,并输出中心点对应的坐标值;根据中心点坐标值可以确定眼睛区域,得到人眼图像;然后将人眼图像输入到1个用于判断眼睛睁闭状态的分类网络,得到眼睛的睁闭状态.试验结果表明:本文提出的方法有效可行,眼睛定位的准确率可达96%,状态分类准确率可达97.07%.相比传统方法,该方法具有较好的鲁棒性和应用前景.A new method based on convolution neural network(CNN)for eye state detection is proposed in this paper.Firstly,a convolution neural network is trained to detect the eye center point.The network can quickly detect the eye center point from the facial image,and get the specific coordinate values.With the coordinate of the eye center point,the eye area can be determined,and the eye images are obtained.Then the eye images are sent to a classified network,which can determine whether the eye are open or close.Experimental results show that the proposed method is effective.The accuracy of eye location is 96%and the accuracy of state classification reaches 97.07%.Compared with the traditional method,this method is more robust and with a better application prospect.

关 键 词：眼睛状态检测 人眼定位 卷积神经网络 关键点检测 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]