A biometric modality based on the seismocardiogram (SCG)

In this paper, a novel biometric modality is proposed based on the seismocardiogram (SCG). The SCG represents the mechanical activities generated by the heart, as measured using accelerometer sensors, and has been found to be an important biomarker for cardiac diseases. This paper explores the biometric potential of the SCG. To this end, an SCG-biometric verification system is designed, focusing mainly on the feature extraction and dimensionality reduction methods: the former is based on autocorrelation, while the latter is based on principal component analysis (PCA). To improve the system accuracy and robustness, a data fusion approach is also implemented, by combining the tri-axial SCG signals. The obtained results indicate that the SCG is a promising physiological signal for biometric applications.