Photoplethysmogram signal based biometric recognition using linear discriminant classifier

Abstract

A preliminary study on photoplethysmogram (PPG) based biometry system is presented here. PPG is a physiological signal related to cardiac output and blood flow saturation in body. Recently it is reported that being an automatic physiological phenomenon PPG and other biosignals can be used as biometric parameters for human authentication. In this work, 12 number of features are extracted from filtered PPG and its derivatives and Linear Discriminant Analysis (LDA) is used for classification over the statistical parameters extracted from the feature set. 100% accuracy is achieved for 15 number of data captured using Biopac MP 45.