基于参数化双谱分析的超宽带生物雷达人体目标非接触检测  

Non-contact Detection of Human Targets via UWB Bioradar Based on Parametric Bispectrum Analysis

作　　者：刘淼[1] 吕昊[1] 李盛[1] 荆西京[1] 王健琪[1] 

机构地区：[1]第四军医大学生物医学工程学院,陕西西安710032

出　　处：《中国医疗设备》2013年第11期15-18,163,共5页China Medical Devices

基　　金：国家自然科学基金项目(61201382);陕西省国际合作重点项目(2013KW30-03);陕西省自然科学基础研究计划项目(2012JQ8022)

摘　　要：超宽带(Ultra Wideband,UWB)生物雷达广泛应用于生命体探测等领域,而人体生命体征信息的非接触检测是其需要解决的最主要问题。本文利用参数化双谱估计算法分析UWB生物雷达回波信号,以非高斯白噪声激励的AR(Auto Regression)参数模型对信号进行建模,并用带约束的三阶平均法对模型的参数进行估计。在此基础上,提出了一种新的基于UWB的人体目标非接触检测算法,并将该算法与现行的基于功率谱密度的检测算法进行对比。实验结果表明,有人位置点信号和无人位置点信号在双谱域表现出明显的差别特性,新算法能更好地实现人体目标的非接触检测。Ultra Wideband （UWB） radar has been widely used in many areas such as life detection and so on. One of the most concerned problems should be solved through UWB radar is the non-contact detection of vital signs information for human. In this paper, a parametric bispectrum estimation algorithm was applied to analyze the UWB echoes and an autoregressive （AR） model driven by non- Gaussian white noise was applied to finish the signal modeling. Then estimate the parameters with the constrained third order mean （CTOM） method. Thus, a novel detection algorithm for human targets based on the UWB radar was developed. Then compare this algorithm with a detection algorithm based on power spectrum density （PSD）. Comparison results showed that there was significant difference between signals of positions with human targets and signals of positions without human targets within bispectrum domain, which indicated that new algorithm can implement the non-contact detection for human targets effectively.

关 键 词：目标探测 UWB雷达 非接触 AR模型 双谱分析法 

分 类 号：R318.04[医药卫生—生物医学工程] R197.39[医药卫生—基础医学]