| 53 | 0 | 68 |
| 下载次数 | 被引频次 | 阅读次数 |
人群拥挤检测对于城市公共安全至关重要。利用城市公共场所现有Wi-Fi信号，应用ESP32芯片采集Wi-Fi信号的CSI数据，采用长短期记忆（Long Short-Term Memory,LSTM）网络建立人群拥挤检测模型，可以实现人群拥挤检测。结果表明：采用CSI数据的均值和方差作为LSTM网络的输入特征指标，模型具有更好的性能；正确率、精确率、召回率和F1评分指标最低值均超过70%；卡尔曼滤波可以将模型性能提高约10%；采用较长的输入序列具有更好的模型性能。与现有主流基于视频图像的人群拥挤检测技术相比，本系统实施成本低、覆盖面广。
Abstract：Crowd identification is vital for urban public safety. By leveraging existing Wi-Fi signals in public places and employing ESP32 chip to collect CSI data, a crowd identification model is established using Long Short-Term Memory(LSTM) networks. Results demonstrate that using the mean and variance of CSI data as input features for the LSTM network yields superior performance; accuracy,precision, recall and F1 score all exceed 70%; Kalman filtering further enhances model performance by approximately 10%; while longer input sequences improve model accuracy. Compared to existing video-based crowd identification technologies, this method offers lower implementation costs and broader coverage.
[1]梁静,杨君涛.大型社会活动安全风险指标体系构建的思考[J].现代职业安全,2015(4):71-73.
[2]卢文刚,蔡裕岚.城市大型群众性活动应急管理研究——以上海外滩“12.31”特大踩踏事件为例[J].城市发展研究,2015,22(5):118-124.
[3]许敏,胡滨.密集场景下的人群拥挤检测研究综述[J].计算机工程,2026,52(3):79-96.
[4]MARANA A N, VELASTIN S A, COSTA L F, et al. Automatic estimation of crowd density using texture[J]. Safety Science, 1998, 28(3):165-175.
[5]Marana A N , Costa L D F , Lotufo R A ,et al.Estimating crowd density with Minkowski fractal dimension[J].IEEE, 2002.
[6]Rahmalan H , Nixon M S , Carter J N.On Crowd Density Estimation for Surveillance[J]. 2006:540-545.
[7]Krausz B , Bauckhage C.Analyzing pedestrian behavior in crowds for automatic detection of congestions[J].IEEE, 2011.
[8]Krausz B,Bauckhage C.Loveparade 2010:Automatic video analysis of a crowd disaster[J].Computer Vision and Image Understanding,2011,116(3):307-319.
[9]HUANG L,CHEN T,WANG Y,et al.Congestion detection of pedestrians using the velocity entropy:A case study of Love Parade 2010 disaster[J].Physica A:Statistical Mechanics and its Applications,2015,440:200-209.
[10]Coianiz,Tarcisio.Estimating the crowding level with a neuro-fuzzy classifier[J].Journal of Electronic Imaging, 1997, 6(3):319.
[11]LO B P L, Velastin S A. Automatic congestion detection system for underground platforms[J]. Proceedings of 2001 International Symposium on Intelligent Multimedia, Video and Speech Processing, ISIMP2001, 2001:158-161.
[12]JEONG J, CHOI J, JO D U, et al. Congestion-Aware Bayesian Loss for Crowd Counting[J]. IEEE Access, 2022, 10:8462-8473.
[13]DUAN J, LIU H, GONG W, et al. Crowd Evacuation Under Real Data:A Crowd Congestion Control Method Based on Sensors and Knowledge Graph[J]. IEEE Sensors Journal, 2023, 23(8):8923-8931.
[14]SALEH B,DAUD S K,EMAD F,et al.Deep learning framework for congestion detection at public places via learning from synthetic data[J].Journal of King Saud University-Computer and Information Sciences,2023,35(1):102-114.
[15]ZHENG Y,ZHOU Z M,LIU Y H.From RSSI to CSI:Indoor Localization via Channel Response[J]. ACM computing surveys, 2014, 46(2):25.1-25.32.
[16]周健.无线感知网络中基于CSI的室内入侵检测与行为识别研究[D].南京:南京邮电大学,2018.
[17]朱海,肖甫,孙力娟,等.基于信道状态信息的WiFi环境感知技术[J].南京邮电大学学报(自然科学版),2016,36(1):94-103,110.
[18]杨寿贵,杨越,刘阳,等.基于ESP32的室内入侵检测系统设计与实现[J].通信与信息技术,2024(4):1-6.
[19]Halperirr D, Hu W, Sheth A, et al. Tool release:gathering802.11n traces with channel state information[J]. ACM Sigcomm Computer Communication Review, 2011, 41(1):53-53.
[20]Xie Y , Li Z , Li M.Precise Power Delay Profiling with Commodity Wi-Fi[J].IEEE Transactions on Mobile Computing, 2015.
[21]Gringoli F , Schulz M , Link J ,et al.Free Your CSI:A Channel State Information Extraction Platform For Modern Wi-Fi Chipsets[C]//the13th International Workshop.2019:21-28.
基本信息:
中图分类号:X91;TN92;TP18
引用信息:
[1]陶琪,彭萧洋,王星亿,等.基于Wi-Fi信号探测的人群拥挤检测[J].通信与信息技术,2026,No.281(03):12-17.
基金信息:
室内空间布局优化与安全保障四川省高校重点实验室（项目编号：2023SNKJ-02）; 多维数据感知与智能信息处理达州市重点实验室（项目编号：DWSJ2411）; 成都师范学院（项目编号：2024XJSXJS1;2024XJCXXSY9;CSTJZX2569;2025xjcxsy02;S202514389127）基金; 2025年教育部产学合作项目资助
2026-03-16
2026
2026-03-25
2026
1
2026-04-09
2026-04-09
2026-04-09