用于身份识别的ECG特征筛选策略研究  

Research on ECG feature selection strategy for identity recognition

作　　者：陈彦会[1] 郑刚[1] CHEN Yan-hui ZHENG Gang(School of Computer and Communication Engineering, Tianjin University of Technology, Tianjin 300384, Chin)

机构地区：[1]天津理工大学计算机与通信工程学院,天津300384

出　　处：《天津理工大学学报》2017年第4期57-61,共5页Journal of Tianjin University of Technology

基　　金：天津市自然科学基金(16JCYBJC15300;15JCYBJC15800)

摘　　要：为了解决基于心电信号(ECG,Electrocardiogram)中基准点特征适用度的问题,在理想化的特征计算结果的基础上,本文提出引入由于计算机识别心电波形基准点所带来的误差,从实际应用角度选择出在身份识别中具有应用价值的特征.研究首先采用小波变换处理心电信号,将心电波形的基准点与不同阶小波系数相对应,自动检测基准点并同时获得检测准确率.然后,设计并实现了由基准点检测准确率估算基准点构成特征的准确率策略,进一步利用逐步判别法进行特征选择,最后,获得特征在用于身份识别中的筛选结果排序.研究采用PTB心电数据库(PhysikalischTechnische Bundesanstalt,德国国家计量科学院心电数据库)作为实验数据,在60人的数据集上进行身份识别,获得98.7%的准确率.In order to solve the problem that the applicability of the fiducial points features based on the ECG signals （ECG, Electrocardiogram）, based on the ideal calculation results of the features, this paper introduces the computer recognition error of ECG fiducial points, selecting features which is useful in the identification of from the practical point of view. Firstly, wavelet transform is used to deal with the ECG signals, and the fiducial points of the ECG waveform and the different wavelet coefficients are corresponding, automatic detection of fiducial points and obtaining the accuracy of detection. Then, the design and implementation of a fiducial point detection accuracy to estimate fiducial points features of the accuracy. Further, stepwise discriminant analysis was used to feature selection. At last, getting the sorting results of features in the identification. Study on the PTB ECG database （Physikalisch-Technische Bundesanstalt, the German National Institute of Metrology Database ） as experimental data, were identified in 60 of the data set, the accuracy rate was 98.7%.

关 键 词：心电信号 身份识别 逐步判别 连续小波变换 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]