Department of Computer Science
Università degli Studi di Milano
Recent studies in biometrics focus on one dimensional
physiological signals commonly acquired in medical
applications, like electrocardiogram (ECG), electroencephalograms
(EEG), phonocardiogram (PCG), and photoplethysmogram
(PPG). In this context, an important application is in
continuous authentication scenarios since physiological signals
are frequently captured for long time periods in order to monitor
the health status of the patients.
With respect to other physiological signals, PPG data present
some advantages since they can be captured using widely diffused,
comfortable, and low-cost sensors.
We have performed a feasibility study on continuous
authentication techniques based on PPG signals. We have realized a biometric
recognition method based on a correlation approach. The accuracy of this method have been evaluated on different datasets describing signals of variable time duration.
Finally, the performance of continuous enrollment strategies have been
The obtained results suggest that PPG signals present sufficient
discriminability to be used in biometric applications that do not
require very high accuracy. Moreover, the use of continuous
enrollment strategies can improve the performance of continuous
Acquisition of a photoplethysmogram (PPG) using a pulse oximeter attached to the fingertip
Example of matching between two templates