QRS complex based human identification

In the last years, the electrocardiogram signal has become an important biometric modality due essentially to the physiological or/and behavioral characteristics variation of the heart among different individuals. The aim of this paper is to present a human identification approach using some time and frequency features of the QRS complex of the ECG signal. These features are extracted from a fractional order model of the frequency content of the QRS complex besides of its temporal area. The K-Nearest Neighbors (KNN) classifier is used for the human identification through the proposed clustering features. Series of tests have been performed to evaluate the proposed identification algorithm using 20 subjects from the MIT-BIH arrhythmia database.