Year: 2018
Type: dissertation
Abstract: This thesis presents a new biometric modality called Photoplethysmograph (PPG) for biometric authentication. PPG records the blood volume change from any part of the body with a combination of a light emitting diode and a photo diode. Therefore, it is the easiest to acquire amongst all medical biometric traits. Although PPG is a time-dependent physiological signal and alters with physical and psychological stress, most of the prior art has neglected these factors. This work presents the ‘BioSec.Lab PPG Database’ and establishes the PPG as biometrics through a comprehensive analysis on its distinctiveness and robustness under exercise, emotions, time-lapse, and noisy conditions. Performances under all variations mentioned above are assessed using various methodologies. Encouraging Equal Error Rates (EER) between 0.00%-36.09% were achieved with relaxed states and under different variations from the three databases.
Source: TSpace (University of Toronto)
Author: Umang Yadav
Language: English
Subfield: Signal Processing
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2908342352