ECG bio-identification using Fréchet classifiers: A proposed methodology based on modeling the dynamic change of the ECG features

Recently, the use of electrocardiogram (ECG) for human identification has attracted great attention. Generally, most existing ECG based biometric systems relay on extracting the static fiducial or non-fiducial features of the cardiac signal. However, the recorded ECG data is more likely to be different whenever it is measured. Such problem can be addressed by utilizing the dynamic change in ECG features. This paper proposes a new methodology …