《计算机应用研究》|Application Research of Computers

基于修正ASM的驾驶员警惕性识别方法研究

Research on recognition method of driver vigilance based on modified ASM

免费全文下载 (已被下载 次)  
获取PDF全文
作者 袁小平,孙辉,王岗
机构 中国矿业大学 信息与控制工程学院,江苏 徐州 221008
统计 摘要被查看 次,已被下载
文章编号 1001-3695(2020)07-057-2192-04
DOI 10.19734/j.issn.1001-3695.2018.12.0919
摘要 针对驾驶员警惕性研究中分析注意力程度的重要性,提出基于修正主动形状模型(ASM)的驾驶员警惕性识别方法。首先建立包含26个特征点的人脸ASM,其次结合面部结构约束构建了双眼平均合成精确滤波器(ASEF),并通过旋转进一步增强鲁棒性,然后用改进ASEF修正人脸ASM;采用左右瞳孔和鼻子特征点建立三角形视线模型,并分析驾驶员注意力程度,利用左右眼角特征点距离对眼睛闭合程度进行归一化,最后利用支持向量机(SVM)分类得到警惕性程度。利用Visual Studio 2017平台进行实验,结果显示,改进ASEF滤波器的准确率达到95.16%,SVM对警惕性程度的分类准确率达到93.8%,每帧平均耗时49.13 ms,表明提出的方法能够有效地识别驾驶员的注意力程度以及警惕性程度。
关键词 警惕性; 注意力程度; 人脸ASM; 改进ASEF; SVM
基金项目 国家科技支撑计划资助项目(2013BAK06B08)
江苏省自然科学基金资助项目(BK20170278)
本文URL http://www.arocmag.com/article/01-2020-07-057.html
英文标题 Research on recognition method of driver vigilance based on modified ASM
作者英文名 Yuan Xiaoping, Sun Hui, Wang Gang
机构英文名 School of Information & Electric Engineering,China University of Mining & Technology,Xuzhou Jiangsu 221008,China
英文摘要 Aiming at the importance of analysis driver's attention level in driver vigilance research, this paper proposed a recognition method of driver vigilance based on modified ASM. Firstly, this paper built a face ASM with 26 feature points. Secondly, it introduced face structure constraint into the ASEF to construct binocular filters, and enhanced the robustness by rotating the filters. Then the improved ASEF modified the face ASM during matching. Thirdly, this paper established a triangle model of sight through binocular and nose feature points from the modified ASM, analyzed the driver's attention level with the model, and normalized eye closure degree by the distance of eye corners. Finally, it applied SVM to classify the different state of driver. Taking experiment on Visual Studio 2017 platform, result shows that the accuracy of the improved ASEF reaches 95.16%, the classification accuracy for driver vigilance reaches 93.8%, and the average time consumed per frame is 49.12 ms, which indicates that the proposed new method can efficiently recognize attention level and the state of driver vigilance.
英文关键词 vigilance; attention level; face active shape model(ASM); improved average of synthetic exact filters(ASEF); support vector machine(SVM)
参考文献 查看稿件参考文献
 
收稿日期 2018/12/13
修回日期 2019/1/29
页码 2192-2195,2201
中图分类号 TP391.41
文献标志码 A