传感技术学报2025，Vol.38Issue(4)：689-694,6.DOI:10.3969/j.issn.1004-1699.2025.04.016
基于毫米波雷达的行为监测研究
Research on Behavior Monitoring Based on Millimeter Wave Radar
摘要
Abstract
With the rapid growth of population,public safety and personal health have become the focus of social management,so intelli-gent behavior monitoring in public places has become very necessary.However,most of the current solutions start from the field of vision and WiFi,and have certain limitations in privacy and measurement accuracy,which can not be better applied in public places.High-reso-lution millimeter wave radar is used to realize behavior monitoring in public space.Firstly,human pose estimation is realized based on ra-dar point cloud data.To compensate for radar point cloud sparseness,PointNet is used as the backbon and self-attention mechanism is combined to build an attitude estimation network to recover human pose from each frame data.Finally,the behavior is identified by tempo-ral convolutional network(TCN).The experimental results show that the average error of key point prediction of human pose estimation is about 2.1 cm,and the accuracy of behavior recognition based on TCN network can reach 93.8％,which has good application value.关键词
毫米波雷达/点云/姿态估计/TCN/行为监测Key words
millimeter wave radar/point cloud/attitude estimation/TCN/behavior monitoring分类
电子信息工程引用本文复制引用
周卫元,张继远,周志湖,徐瑞吉,程燃,毛科技,赵永标..基于毫米波雷达的行为监测研究[J].传感技术学报,2025,38(4):689-694,6.基金项目
浙江省基础公益研究计划项目(LTGG23F020002,LGG22F020014) （LTGG23F020002,LGG22F020014）
国家自然科学基金项目(62072410) （62072410）
浙江省教育厅一般科研项目(2024) （2024）
