基于微多普勒特征的人体步态参数估计  

Human Gait Parameter Estimation Based on micro-Doppler Feature

作　　者：张翼[1] 邱兆坤[2] 朱玉鹏[3,4] 黎湘[2] 

机构地区：[1]北京系统工程研究所 [2]国防科技大学空间电子信息技术研究所 [3]北京市北方电子设备研究所 [4]中国人民解放军61906部队

出　　处：《信号处理》2010年第6期917-922,共6页Journal of Signal Processing

摘　　要：基于雷达的人体微动研究是近年发展起来的新技术,通过信号处理手段可获取人体走动时手和腿的摆动激励的多普勒频移,用于人体步态探测与识别。本文首先分析了人体行走对雷达波的调制,介绍了一套用于人体运动特性测量的X波段连续波雷达系统及测量实验,并对人体行走实测数据中微多普勒特征进行分析,然后深入研究了P.van Dorp等人提出的基于微多普勒特征的人体行走实测数据特征提取和参数估计的方法,并将之应用于本雷达系统的测量结果中,最后给出了仿真结果。Human micro-motion study based on radar is a new technique developed in recent years.Through signal disposal measures,Doppler shift of human arms and legs during walking can be gained so as that human pace may be detected and recognized.The paper first analyzed the modulating of human walking on radar wave,introduced the X-band continuous wave radar system used on human motion feature measurement and relative measurement experiment,and analyzed the micro-doppler feature of human walking through the field-measured data.The paper deeply studied the methods of feature parameter extraction and human pace parameter estimation raised by P.van Dorp,and used it in the measurement result of this radar system.Finally the paper gave the simulation result.

关 键 词：人体微动 微多普勒 参数估计 人体步态特征 

分 类 号：TN95[电子电信—信号与信息处理]