• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

J4 ›› 2015, Vol. 37 ›› Issue (12): 2379-2385.

• 论文 • 上一篇    下一篇

基于Kinect的疲劳驾驶综合监测系统研究

蔡菁,程雷,王红霞   

  1. (武汉理工大学计算机科学与技术学院,湖北 武汉 430063)
  • 收稿日期:2015-08-12 修回日期:2015-10-17 出版日期:2015-12-25 发布日期:2015-12-25
  • 基金资助:

    湖北省自然科学基金资助项目(2013CFB351);武汉理工大学校自主创新基金资助项目(2014IV105)

A driver fatigue state monitoring
system  based on Kinect   

CAI Jing,CHENG Lei,WANG Hongxia   

  1. (School of Computer Science and Technology,Wuhan University of Technology,Wuhan 430063,China)
  • Received:2015-08-12 Revised:2015-10-17 Online:2015-12-25 Published:2015-12-25

摘要:

基于Kinect传感器研究设计一种驾驶员疲劳状态综合监测系统,通过对Kinect红外图像数据的预处理,减弱了夜晚光照不足的影响;进而利用Kinect提供的人脸识别功能获取驾驶员头部、嘴部、眼部等部位的特征信息,并利用RBF神经网络进行信息融合,分级判断驾驶员的疲劳状态;同时利用滑动平均法及数据库技术,使疲劳状态监测更加准确可靠。模拟实验结果表明,本系统在白天甚至夜晚都能较有效地监测驾驶员疲劳状态。

关键词: 驾驶监测, 疲劳监测, 红外图像, RBF, 信息融合

Abstract:

We present a driver fatigue state monitoring system for based on Kinect. Preprocessing the Kinect infrared image data can weaken the effect of light deficiency at night. Moreover, the processed images and the face recognition provided by Kinect are used to obtain the features of adriver’s head, mouth and eyes, which are further fused via RBF neural network to judge the fatigue state of drivers on different levels. The identification of fatigue state becomes more accurate and reliable with the help of the sliding average method and database technique. Simulation results show that the proposed system can effectively monitor the fatigue state of drivers at day time, or even at night.

Key words: driving monitoring;fatigue monitoring;infrared image;RBF;information fusion