摘要
为了促进智慧城市的发展,提升小区门禁系统的智能化服务,设计了一种基于人脸识别的智慧小区门禁系统。讨论了系统的总体架构,各模块功能作用,控制流程以及软件设计等,系统分为用户操作端部分、前端门禁机部分以及后台识别比对部分,主要包括用户信息注册模块、前端抓拍识别模块、数据传输模块、后台识别比对模块、智能开锁模块以及中心管控模块。人脸检测部分采用改进的AdaBoost算法,提高了多姿态下人脸检测的鲁棒性,后台比对部分采用多层卷积神经网络的方法对人脸进行比对。经过现场的实际运行,此系统方案具有良好的可靠性和稳定性。
In order to promote the development of smart city and enhance the intelligent service of community access control system,a smart community access control system based on face recognition is designed.The overall architecture of the system,the function of each module,the control flow and the software design are discussed.The system is divided into the user operation part,the front-end access control part and the background recognition comparison part,including the user information registration module and the front-end capture recognition module.Data transmission module,background recognition comparison module,intelligent unlocking module and central management module.The face detection part adopts the improved AdaBoost algorithm to improve the robustness of face detection under multi-pose.The background comparison method uses multi-layer convolutional neural network to compare faces.After the actual operation on site,this system solution has good reliability and stability.
作者
阮有兵
徐海黎
万旭
沈标
RUAN Youbing;XU Haili;WAN Xu;SHEN Biao(School of Mechanical Engineering, Nantong University, Nantong 226019, China;Nanjing Lantai Traffic Establishment Co., LTD., Nanjing 210019, China)
出处
《电视技术》
2019年第1期90-93,120,共5页
Video Engineering
关键词
人脸识别
门禁
face recognition
access control
