基于面部识别的管制员疲劳风险评估关键技术研究  

Face Recognition for Fatigue Risk Assessment of Air Traffic Controllers

作　　者：孙瑞山[1] 史增鹏 王俊虎[1] 

机构地区：[1]中国民航大学民航安全科学研究所,天津300300

出　　处：《交通信息与安全》2014年第1期1-4,共4页Journal of Transport Information and Safety

基　　金：中国民航局科技项目(重大专项)(批准号:MHRD2011238);国家重点基础研究发展计划(973计划)项目(批准号:2010CB734105)资助

摘　　要：人为因素在民航事故发生原因中所占比例逐渐增加,因此,高效检测管制员的身份和疲劳状况,对保证飞行正常及安全风险的评估具有重要的意义。基于面部识别技术,采用OpenCV开发平台,综合图象二值化和模板匹配技术,对管制员的身份及疲劳状况进行识别,取得比较实际的检测效果,疲劳检测成功率达80%,对管制员这类工作岗位的风险管理具有一定的参考价值。The major responsibility of air traffic controllers is dispatching and routing air traffic to prevent aircraft collisions ,which demands vigor and highly concentrated attention .Therefore ,the efficient detection of the controllers＆#39;i-dentity and their fatigue conditions plays a significant role in guaranteeing normal flight and safety performance evalua -tion .Based on the facial recognition technology ,this paper takes advantage of the OpenCV development platform and im-age binarization method as well as template matching technology to detect the controllers ＆#39;identity and their fatigue condi-tions .Comparatively detection effect has been obtained with a fatigue testing rate of 80% ,which provides the controllers of this risk management work with some references .

关 键 词：管制员疲劳检测系统 眼部检测 人脸识别 OPENCV 

分 类 号：TP317.4[自动化与计算机技术—计算机软件与理论] X951[自动化与计算机技术—计算机科学与技术]