基于大数据的人脸识别方法  

Method of face recognition based on big data

作　　者：王彦秋 冯英伟 WANG Yanqiu;FENG Yingwei(Modern Education Technical Center,Eastern Liaoning University,Dandong 118001,China;Modern Education Technical Center,Hebei University of Architecture,Zhangjiakou 075000,China)

机构地区：[1]辽东学院现代教育技术中心,辽宁丹东118001 [2]河北建筑工程学院现代教育技术中心,河北张家口075000

出　　处：《现代电子技术》2021年第7期87-90,共4页Modern Electronics Technique

基　　金：2019年度河北省高等学校科学研究课题(QN2019180)。

摘　　要：针对目前人脸识别存在的主要安全隐患,包括验证数据库安全问题及假体攻击问题,提出基于大数据的人脸识别算法。设计了大数据人脸识别算法的系统架构,将系统分为前端采集终端、云端数据处理及数据通信三大部分。在该架构基础上,设计大数据人脸识别算法的实现流程,前端采用PFLD算法进行人脸特征点提取,通过网络将数据传输至云端存储,并在云端进行计算识别,充分利用云服务器的计算与存储优势。最后进行了实验与仿真,结果表明,基于该架构的大数据人脸识别算法在网络传输为理想状态下,能大大提高人脸识别的精度,提高了验证数据库的安全性。In view of the main potential risks of face recognition including the security of database verification and prosthesis attack,a face recognition algorithm based on big data is proposed.The system architecture of the algorithm is designed,which is divided into front-end acquisition terminal,cloud data processing and data communication.On the basis of this architecture,the implementation process of the algorithm is designed.At the architecture front-end,PFLD(practical facial landmark detector)is used to extract facial feature points.The extracted feature data is transmitted to the cloud terminal by the network for storage,and then for calculation and recognition,which makes full use of the computing and storage advantages of cloud server.The experiments and simulations were carried out.The results show that the proposed algorithm can greatly improve the accuracy of face recognition and the security of the verification database under the ideal state of network transmission.

关 键 词：人脸识别 大数据 系统架构设计 人脸特征提取 算法流程设计 云计算 

分 类 号：TN911.73-34[电子电信—通信与信息系统] TP393.1[电子电信—信息与通信工程]