部分遮挡人脸识别在门禁管理中的应用  

Application of partially occluded face recognition in access control management

作　　者：成洪豪 王学贺[1] CHENG Honghao;WANG Xuehe(Division of Computer Science,Heze Medical College,Heze 274000,Shandong,China)

机构地区：[1]菏泽医学专科学校计算机教研室,山东菏泽274000

出　　处：《智能计算机与应用》2024年第9期70-75,共6页Intelligent Computer and Applications

摘　　要：近年来,由于新冠疫情等公共卫生事件的影响,佩戴口罩成了日常生活的新常态,这也给传统的人脸识别技术带来了巨大挑战。在高校实验室等特殊场景下,由于实验环境的敏感性和安全要求,有必要对实验人员进行身份识别和行为监控,因此亟需一种能够在口罩遮挡情况下准确识别人脸的解决方案。本文设计了一种基于MTCNN和改进FaceNet的佩戴口罩人脸识别系统,并将其应用于高校实验室场景。人脸识别结果将被传回客户端,同时系统还会在Kurento服务器中存储一段时间的视频片段,可以有效支持实验室安全管理和行为审计。In recent years,due to the impact of public health events such as the COVID-19 pandemic,wearing masks has become the new normal in daily life.This has posed significant challenges to traditional facial recognition technologies.In special scenarios such as university laboratories,where sensitivity and safety requirements demand the identification and monitoring of individuals who often need to wear masks,there is an urgent need for a solution that can accurately recognize faces even when they are partially obscured by masks.This paper proposes a mask-wearing face recognition system based on MTCNN(Multi-Task Cascaded Convolutional Networks)and FaceNet,which is applied in the context of university laboratories.The results of face recognition are transmitted back to the client,and the system also stores video clips for a certain period in the Kurento server,effectively supporting laboratory security management and behavioral auditing.The system has been piloted in multiple laboratories and has achieved good application results.It is expected to provide an efficient and accurate personnel management and security solution for special scenarios such as university laboratories.

关 键 词：人脸识别 部分遮挡 口罩识别 多任务级联卷积网络 身份识别 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]