Biometric Identification for Twins Using Photoplethysmogram Signals

This paper focuses on the photoplethysmogram (PPG) based biometric identification for twins as a substitute to the traditional methods being used nowadays. There are many researches and studies about twin recognition where some of them showed positive results. However, PPG based biometric identification for twins is under-researched. Therefore, this issue will be the main focus of our study. A total of nine subjects were used for experimentation purposes. A lowpass filter is used to remove unwanted portion of the signal. Unique features are extracted from these filtered PPG signals. Later, Naïve Bayes (NB), Radial Basis Function (RBF) and Multilayer Perceptron (MLP) are used to classify the subjects using the discriminant features. Based on the experimentation results, classification accuracies of 95.4%, 94.4% and 88.9% were obtained when using NB, RBF and MLP respectively. The result provides an alternative mechanism to detect a person besides using the traditional methods.