摘要
根据江门供电局实际工程项目所提出的智慧工地,针对其中的门禁系统需求,使用OpenCV的接口和多线程技术进行实时图像采集,采用第三方开源的人脸识别库FaceRecognitionDotNet进行设计,在检测上采用了Hog特征,在识别验证采用了Dlib开源的ResNet模型,以此为基础设计智慧工地门禁系统软件,实现施工人员考勤、考勤数据可视化和施工人员数据管理等功能。
According to the smart construction site proposed by the actual engineering project of Jiangmen Power Supply Bureau,in response to the access control system requirements,the OpenCV interface and multithreading technology are used for real-time image acquisition.The Hog feature is used for detection,and the Dlib open source ResNet model is used for recognition and verification.Based on this,the smart site access control system software is designed to realize the functions of construction staff attendance,attendance data visualization and construction staff data management.
作者
邹媛媛
练文广
黄健光
张红阳
周远波
林润钊
蔡庆荣
ZOU Yuanyuan;LIAN Wenguang;HUAN Jianguang;ZHANG Hongyang;ZHOU Yuanbo;LING Runzhao;CAI Qingrong(Wuyi University,Jiangmen 529000,China;Minghao Electric Power Supervision Co.,Ltd.,Jiangmen 529000,China)
出处
《电工技术》
2021年第8期36-39,共4页
Electric Engineering
基金
五邑大学基建工程智慧工地服务科研咨询服务项目(编号HX19083)。
关键词
门禁系统
人脸识别
考勤
access control system
face recognition
attendance
