基于多层神经网络的高校智能安保应用研究  

Research on the Application of Intelligent Security in Universities Based on Multilayer Neural Network

作　　者：李小丽 LI Xiao-li(Quanzhou College of Economics and Trade,Quanzhou Fujian 362000,China)

机构地区：[1]泉州经贸学院,福建泉州362000

出　　处：《河北软件职业技术学院学报》2022年第2期1-3,共3页Journal of Hebei Software Institute

基　　金：福建省教育厅中青年科技项目“基于云计算的远程电气节能智慧控制平台”(JAT191470)。

摘　　要：安保工作在高校日常管理中占有重要地位,传统视频监控主要进行视频录制,作为数据资料进行保存,一般在校园中无安全事件发生时,视频数据基本不会再使用,数据保存量相当大;当出现异常时,人工进行数据检视与排查,效率也很低。智能安保系统采用多层神经网络算法对视频数据进行分割、训练与预排查,并对人体行为进行分析,根据结果预测是否报警,过程无需人工干预,为高校安保工作带来及大便利。The security work occupies an important position in the university management,the traditional video surveillance mainly uses video recording as data to save.Generally,when there is no security event on campus,the video data will not be used again,and the amount of data saved is quite large.However,in case of abnormality,the efficiency of manual data inspection and troubleshooting is very low.The intelligent security system uses multilayer neural network algorithm to segment,train and pre-check the video data,analyze the human behavior,and predict whether to alarm according to the results.There is no need for manual intervention in the process,which brings great convenience to the security work of the universities.

关 键 词：多层神经网络 高校智能安保 视频图像序列 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]