Towards Real Time Implementation of Sparse Representation Classifier (SRC) Based Heartbeat Biometric System

Abstract

Implementation of the heartbeat biometric system consists of four main stages which are heartbeat data acquisition, pre-processing and feature extraction, modeling and classification. In this study a new approach for classification method based on Sparse Representation Classifier (SRC) is proposed. By introducing kernel trick into SRC, the classification performance of the classifier can be further improved by implicitly map features data into a high-dimensional kernel feature space. Based on heart sound data, experimental results have shown a promising performance of KSRC with 85.45 % of accuracy has been achieved and a better performance has been observed by this classifier compared to Support Vector Machines (SVM), SRC and K-Nearest Neighbor (KNN). This achievement has proved the possibility of heartbeat as a biometric trait for human authentication system. Due to this, an extension in term of heartbeat data acquisition toward real time implementation is then proposed in this paper. Here, a wrist-mounted heartbeat sensor to sense the heartbeat signal is designed. This developed sensor is an electrometer which is capable to measure the properties of electrocardiogram (ECG) signal. The developed hardware has also shown its viability toward execution of heartbeat data acquisition in real time.