基于PCA特征和融合特征的ECG身份识别方法

ECG作为一种活体生物特征用于身份识别在国际上引起了广泛重视.针对基于解析特征的ECG身份识别方法对特征点检测精度要求较高的缺点，提出一种仅需R波峰值点检测的ECG身份识别方法，该方法通过有针对性的设定相应阈值，将PCA特征和小波融合特征方法相结合.实验结果表明该方法优于PCA特征方法、波形特征方法和小波特征方法，既减少了特征点检测的复杂性和特征点检测不准确带来的误差，又可获得较高的识别率，是一种实时、高效算法.

A method based on the PCA feature and fusion feature for ECG human identification

As a new biometric for identification, ECG attracted widespread attention in the international community. The method was based on the analytic feature for ECG identification and required high precision for fiducial points detection. To overcome this disadvantage, a method, in which only the peak point of the R-wave detection is needed, was proposed. As for setting the relevant threshold, this method combined the PCA feature method and fusion feature method based on wavelet decomposition. The experiment demonstrates that the method proposed in this paper is better than the PCA feature method, waveform feature method, and wavelet feature method. This method, which not only reduces the complexity and error of the fiducial points detection but also achieves high accuracy, is a realistic and efficient algorithm.