基于多重分形特征的心电身份识别方法研究

Research on ECG Identification Method Based on Multifractal Features

由于人的心脏是一个复杂的非线性混沌系统,利用多重分形分析手段来研究心电(electrocardiosignal,简称ECG)的混沌也是一种趋势.文章利用质量指数、多重分形谱两个概念,定性地分析了心电信号具有多重分形特性.并利用质量指数对称度、多重分形谱宽度、最大\最小概率子集分形维数差、多重分形谱非对称指数作为心跳周期的多重分形特征,结合支持向量机(Support Vector Machine,简称SVM)在3个心电数据库进行仿真测试.测试得到的心跳判别结果通过多数投票规则以判断受测者的身份,身份识别率达到96.67%,是一种可行的识别方法.

Since the human heart is a complex nonlinear chaotic system,it is a trend to study the chaos of ECG by using multifractal analysis.In this paper,two concepts of mass index and multifractal spectrum are used to qualitatively analyze the multifractal characteristics of ECG signal.The quality index's symmetry,the width of multifractal spectrum,the fractal dimension difference in max/min probability subset and the asymmetric index of multifractal spectrum are used as the multifractal characteristics of single cardiac cycle.And test in the three ECG databases combined with support vector machine(SVM).The heartbeat discrimination results obtained from the test are used to determine the participants'identity by majority voting rules.At last the identity recognition rate reaches to 96.67%.It is a feasible method for identification.