Biometric Clustering of ECG using Wave Peaks

Abstract

The use of ECG signals for biometric recognition is in the focus of scientific research. For each person electrocardiogram which contain specific biometric characteristics can be recorded making it suitable for biometric application. A comparison of features in terms of potential person identification, i.e. clustering needs, is made, where amplitude characteristics are extracted in time domain. This paper analyzes data from the Physionet ECG-ID database, and show promising results for future ECG based considerations.