Investigation: Use of electrocardiogram as a biometric over longer time period

ECG based biometric is a recent topic of research. Many researchers have proved that ECG can be used as a biometric for person identification. ECG is time variant signal. This paper presents an analysis on the possibility to use electrocardiogram as a biometric over a longer time period. Data used during the experimentation is recorded from two different subjects, within thirty six months. Lead II signal is used for the analysis purpose. One sample from the recording of each of the two subjects is considered as an enrolled database and the remaining samples are used for the testing purpose. Analysis is made using discrete wavelet transform, which showed the accuracy of more than 91% to conclude that the ECG signals are sufficiently distinct for an individual. ECG as a biometric has an advantage as it checks the aliveness of an individual and the possibility of falsification can be removed.