The heart rate detection algorithm through the voice signal by combining short-term Fourier transform (STFT) and principal component analysis (PCA) methods is presented for non-contact physiological signal acquisition applications. The voice signal pa...
The heart rate detection algorithm through the voice signal by combining short-term Fourier transform (STFT) and principal component analysis (PCA) methods is presented for non-contact physiological signal acquisition applications. The voice signal passes through the vocal cords consists of muscles, so the heart rate signal is modulated and output. It is possible to improve the inconvenience use of the existing contact type bio-signal detecting device using ECG and PPG. The STFT extracts the signals at various frequencies as it changes over time. Then the formants, ranges of frequencies in which there is a relative maximum in the signal spectrum, is detected to apply them to the PCA algorithm. The PCA separates heart rate signals from the unwanted interferers to calculate the heart rate. The proposed algorithm is verified by comparing it with the PPG-based heart rate detection, resulting in less than 10% error.