当前位置:主页 > 科技论文 > 信息工程论文 >

采用同步挤压小波变换的人体运动姿态分析

发布时间:2018-04-04 09:59

  本文选题:同步挤压小波 切入点:雷达回波 出处:《西安交通大学学报》2017年12期


【摘要】:针对人体运动的雷达回波信号特征复杂、不同运动姿态微多普勒频率差异小、难以区分精细特征的问题,提出了一种采用参数可调的同步挤压小波变换(SSTAP)的人体运动姿态分析方法。首先根据实测人体运动数据构建人体运动模型及其雷达回波模型;然后利用SSTAP方法对人体运动模型雷达回波信号进行分解,获得人体各主要部位的时频特征;再通过调整同步挤压小波变换的2个参数获得人体整体回波信号的具有最佳时频分辨率的时频特征,进一步与各部位的人体时频特征比较获得了人体运动姿态的信息。实验结果表明,相比广义S变换(GST)、小波变换(WT)等时频分析方法,基于参数可调的同步挤压小波的人体微多普勒分析结果更加清晰精细,更能反映人体微运动的特征,其微多普勒频率的分辨率比GST、WT分别提高了17%和14%。
[Abstract]:The characteristics of radar echo signal of human body motion are complex, the difference of micro-Doppler frequency of different motion attitude is small, so it is difficult to distinguish fine features.A method of human motion attitude analysis using synchronous extrusion wavelet transform (SSTAP) with adjustable parameters is proposed.Firstly, the human motion model and radar echo model are constructed according to the measured human motion data, then the radar echo signal of human motion model is decomposed by SSTAP method, and the time-frequency characteristics of the main parts of human body are obtained.Then, by adjusting the two parameters of synchronous extrusion wavelet transform, the time-frequency features of the human whole echo signal with the best time-frequency resolution are obtained, and the human motion attitude information is further obtained compared with the human body time-frequency characteristics of various parts.The experimental results show that compared with generalized S-transform (GSTT), wavelet transform (WTT) and other time-frequency analysis methods, the results of human micro-Doppler analysis based on synchronous extrusion wavelet with adjustable parameters are clearer and finer, and can reflect the characteristics of human micro-motion more clearly.The resolution of microDoppler frequency is 17% and 14% higher than that of GSTT WT, respectively.
【作者单位】: 西安理工大学自动化与信息工程学院;后勤工程学院国防建筑规划与环境工程系;
【基金】:国家自然科学基金重大研究计划资助项目(41390454)
【分类号】:TN957.51

【相似文献】

相关期刊论文 前10条

1 王波,

本文编号:1709461


资料下载
论文发表

本文链接:https://www.wllwen.com/kejilunwen/xinxigongchenglunwen/1709461.html


Copyright(c)文论论文网All Rights Reserved | 网站地图 |

版权申明:资料由用户ad69a***提供,本站仅收录摘要或目录,作者需要删除请E-mail邮箱bigeng88@qq.com