基于深度学习的人脸识别课堂考勤系统的实现  

Implementation of face recognition classroom attendance system based on deep learning

作　　者：邓熠 毕磊 薛甜 范亚江 侯丹 Deng Yi;Bi Lei;Xue Tian;Fan Yajiang;Hou Dan(School of computer science,Xianyang Normal University,Xianyang 712000,China)

机构地区：[1]咸阳师范学院计算机学院,陕西咸阳712000

出　　处：《无线互联科技》2021年第14期50-52,60,共4页Wireless Internet Technology

基　　金：咸阳师范学院大学生创新创业训练计划项目,项目名称:基于人脸识别的课堂考勤系统,项目编号:S202010722014。

摘　　要：文章针对传统考勤效率过低和现有人脸识别技术需要近距离刷脸的弊端,设计了一个基于深度学习的人脸识别课堂实时考勤系统。该系统以树莓派和Intel Movidius神经棒NCS为硬件平台,使用Python和OpenCV实现了基于深度卷积神经网络的人脸识别,并通过设计的考勤系统来保存识别结果和管理考勤数据。这种新型考勤系统在简化考勤流程的同时避免了传统考勤机需要近距离刷脸的弊端,支持远距离多人同时与实时识别,提高了考勤效率。In this paper,a real-time attendance system for face recognition classroom based on depth learning is designed in view of the disadvantages of low efficiency of traditional attendance and the need of close brush face recognition technology.The system uses raspberry pie and Intel Movidius stick NCS as hardware platform,uses Python and OpenCV to realize face recognition based on deep convolution neural network(CNN),and saves recognition results and manages attendance data through the designed attendance system.This new attendance system can simplify the attendance process and avoid the disadvantages of the traditional attendance machine which needs to brush the face at close range,support the recognition of many people at long distance and real time,and improve the efficiency of attendance.

关 键 词：人脸识别 考勤系统 卷积神经网络 结构查询语言 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TP18[自动化与计算机技术—计算机科学与技术]