Abstract:
This paper introduce a new Bipolar Pulse Active
(BPA) feature extraction technique implemented to
electrocardiograms (ECG) for biometric authentication. . The
BPA extracts information from ECG signals and decomposes
them, using a series of harmonically related periodic triangular
waveforms, into a finite set of Pulse Domain features. In this
work, BPA is used to compare the performance of ECG when the
information taken from 3 different locations, namely peaks P to
T, peaks P to R and peaks R to T. The authentication
performance is analysed with and without the use of classifier. In
this work, Linear Discriminant Analysis (LDA) is used a
classifier to evaluate the BPA ECG based features for biometric
authentication.
