Signal processing for heart rate identifcation from radial pulse acceleration measurements

Abstract

This thesis focuses on signal processing methods to identify heart rate from radial pulse acceleration measurements collected by an accelerometer sensor. In the thesis, the theories behind the methods - preprocessing (signal resampling, normalization, and wavelet denoising), recomposition (principal component analysis), and analysis (continuous wavelet transform) - are studied, providing the basis for their implementations. The fnal result is a functional mobile application that can collect radial pulse acceleration measurements using the phone-embedded accelerometer sensor and extract heart-related information.