基于人脸测温技术的通道闸机人员身份识别方法  

Personnel identification method of channel gate based on face temperature measurement technology

作　　者：杨尔欣 刘育平 祝存平 高攀 吴相荣 王星龙 YANG Er-xin;LIU Yu-ping;ZHU Cun-ping;GAO Pan;WU Xiang-rong;WANG Xing-long(State Grid Gansu Electric Power Company Logistics Department,Lanzhou 730000,China;State Grid Siji Feitian(Lanzhou)Cloud Data Technology Co.,Ltd.,Lanzhou 730000,China)

机构地区：[1]国网甘肃省电力公司后勤工作部,兰州730000 [2]国网思极飞天(兰州)云数科技有限公司,兰州730000

出　　处：《信息技术》2023年第10期112-117,共6页Information Technology

摘　　要：以往体温测量方式效率较低,无法满足疫情防控的需求,为了测量流动人员体温,保障公共空间的安全,提出一种基于人脸测温技术的通道闸机人员身份识别方法。利用红外热像仪采集并处理热红外人脸图像;借助人脸测温技术中的黑体模块构建温度——灰度模型,标注超温人员人脸图像;融合Gabor与SVD算法,提取超温人员人脸图像特征向量,构建与集成多分类器,识别超温人员的身份。实验数据显示,在不同特征向量维数下,应用提出方法获得的人脸测温误差较低,人员身份识别率高于最低限值,充分证实该方法的可行性。The efficiency of the former method of temperature measurement is low,which cannot meet the requirment of epidemic prevention.In order to measure the body temperature of floating personnel and ensure the safety of public space,a channel gate personnel identification method based on face temperature measurement technology is proposed.The infrared thermal imager is used to collect and process the thermal infrared face image.With the help of the blackbody module in the face temperature measurement technology,the temperature gray model is constructed,the face image of overtemperature personnel is marked.And,the Gabor and SVD algorithm are fused,the feature vector of the face image of overtemperature personnel is extracted.and multiple classifiers are constructed and integrated to identify the overtemperature personnel.The experiment data show that under different feature vector dimensions,the face temperature measurement obtained by the proposed method has low error,and the personnel identification rate is higher than the minimum limit,which fully proves the feasibility of this method.

关 键 词：人脸测温技术 通道闸机 身份识别 人脸检测 

分 类 号：TN215[电子电信—物理电子学]