步态识别的小波去噪质量评价方法  

Quality evaluation method for wavelet de-noising in gait recognition

作　　者：涂斌斌[1,2] 谷丽华[1] 许会[1] 

机构地区：[1]沈阳工业大学信息科学与工程学院,沈阳110870 [2]沈阳大学信息工程学院,沈阳110044

出　　处：《沈阳工业大学学报》2017年第1期61-66,共6页Journal of Shenyang University of Technology

基　　金：国家自然科学基金资助项目(61372154)

摘　　要：针对步态加速度信号由于采集信号包含大量尖峰或突变等高频噪声,导致在步态分析、周期划分和特征提取方面存在困难的问题,提出了一种利用小波去噪质量评价来对加速度信号小波去噪的最优参数进行选择的新方法.根据质量评价指标的几何和物理意义,使用评价指标值来指导小波去噪最优小波分解与重构层数的参数选择.当评价指标值取最小值时,其对应分解层数为最优分解层数,采用的小波基函数为最优小波基函数.结果表明,本文方法在步态识别领域可以为步态信号小波去噪选择最优小波基函数和分解层数提供理论依据.In order to overcome the difficulties in gait analysis,cycle division and feature extraction for gait acceleration singals due to the fact that the acquired signals contain a large number of higher frequency noises such as peaks or mutation,a new method,where the quality evaluation for the wavelet de-noising was adopted to select the optimal parameters for the wavelet de-noising of acceleration signals,was proposed.According to the geometric and physical meanings of quality evaluation indexes,the paramenter selection for both decomposition and reconsitution levels of optimal wavelet in the wavelet de-noising were guided by the evaluation index value.When the evaluation index took the minimum value,the correspongding decomposition level was optimal,and the used wavelet basis funtion was also optimal.The results show that the proposed method can provid the theoretical basis for the selection of both optimal wavelet basis funtion and decomposition levels for the wavelet de-noising of gait signals in the field of gait recognition.

关 键 词：步态 步态识别 加速度信号 小波去噪 质量评价 小波基函数 分解与重构 最优参数 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]