Real-time continuous identification system using ECG signals

This paper presents a prototype biometric system that per- forms real time human identification using the ECG signal. Cardiac signals have been lately suggested for biometric deployment, since they carry subject specific information and at the same time allow for inherent liveness detection. The HeartID is the first system to continuously authenticate individuals who are conveniently wearing a portable, wireless ECG sensor. The AC/LDA algorithm is used for biometric template design because of its computational and performance advantages. The performance of the prototype testbed was evaluated over 10 subjects at the University of Toronto with very promising results.