Computer Engineering and Applications ›› 2010, Vol. 46 ›› Issue (20): 181-183.DOI: 10.3778/j.issn.1002-8331.2010.20.050

• 人工智能 • Previous Articles     Next Articles

Method of eye location and states analysis

TAN Tai-zhi,LI Ding-lun,LIU Fu-chun   

  1. Faculty of Computer,Guangdong University of Technology,Guangzhou 510006,China
  • Received:2010-04-14 Revised:2010-05-17 Online:2010-07-11 Published:2010-07-11
  • Contact: TAN Tai-zhi



  1. 广东工业大学 计算机学院,广州 510006
  • 通讯作者: 谭台哲

Abstract: Majority of vision-based methods are carried on through the analysis of eye states so as to confirm the states of fatigue,among which the accurate localization of the eyes and the precise analysis of the eye states are the key links.This paper puts forward an efficient eye localization and eye states analysis method,which detects the face under natural environment in advance,and then locates the eyes accurately as well as analyzes the eye states.First of all,an isopleths method is applied to accurately detect the regions of the eyes and preprocess the image,and then the eye states can be determined by using the mathematical morphology method.Experimental results show that the means raised can react the continuously changing states of the eyes with efficiency and high accuracy,in addition it also meets the needs of application requirements of fatigue detection.

Key words: eye localization, isophote curvature, mathematical morphology, eye states, fatigue detection

摘要: 基于计算机视觉处理的方式大多都是通过分析眼睛的状态来判断驾驶员的疲劳状态,其中,眼睛的精确定位和眼睛状态的准确分析是关键环节。提出一种有效的眼睛精确定位和眼睛状态分析方法,通过在自然环境下先检测出人脸,然后进行眼睛的精准定位和眼睛状态分析。首先,使用等亮度线方法进行眼睛区域精确定位和图像预处理,然后使用数学形态学方法分析眼睛的状态。实验结果表明,该方法能很好地反应出眼睛的变化状态,有较高的准确性,能满足疲劳检测的应用要求。

关键词: 眼睛定位, 等亮度线, 数学形态学, 眼睛状态, 疲劳检测

CLC Number: