基于心电信号的身份识别系统研究

Research on Identification System Based on ECG Signal

目的设计一套基于心电信号的身份识别系统,对采集到的心电信号进行算法设计,进而实现身份识别功能。方法首先采集心电数据:一部分是MIT数据库的心电数据,一部分是用便携式心电采集装置采集的心电数据。然后进行心电信号特征提取,在每个心动周期内先找到R波的最高点,然后再提取R波最高点前80个点和后170个点作为处理对象。最后在MATLAB中采用主成分分析(PCA)方法进行数据降维,采用线性判别分析(LDA)方法设计分类器,并进行测试。结果实验结果表明,该心电采集设备可以实时采集心电数据,分类后的数据具有唯一性且分类的正确率达到了90%以上。结论本系统在技术与应用层面上具有一定的创新性,可为社会的安全性提供一定的保障,具有一定应用价值。

Objective To design a set of identification system based on ECG signal and to carry out algorithm design for the ECG signal so as to realize the function of identity recognition. MethodsFirstly, ECG data was collected partly from MIT database of the ECG data and partly from a portable ECG acquisition device. Then the features of ECG signals were extracted and the highest point of each R wave was located. We extracted 80 points before the highest point of the R wave and 170 points after the highest point. Finally, the principal component analysis（PCA） was used to reduce the dimension of data in matrix laboratory（MATLAB）. The lineal discriminant analysis（LDA） was adopted to design a categorizer, which was also tested. Results The experimental result indicated that the use of portable ECG acquisition equipment could collect ECG data in real time. The data after classification was unique and the correct classification rate was above 90%. Conclusion This system has high innovation at the level of technology and application, which could provide certain guarantee for the security of the society, and has certain application value.