A new method for ECG biometric recognition using a hierarchical scheme classifier

Abstract

In this paper, a new method for ECG biometric recognition using a hierarchical scheme classifier is presented. The integral process of the method is introduced, including preprocessing, feature extraction and classification. To achieve a better performance of proposed method, cross-validation is applied to determine the parameters in the classifier. As a result, proposed method offers considerably high recognition rates when tested on MIT-BIH NSRDB. The total heartbeat recognition rate is 97.98%, and window recognition rate and subject recognition rate are both 100%.