One of the most relevant aspects investigated by the scientific community in the healthcare field is the detection of vital parameters, such as the heart rate, through the use of contactless technologies, often considered more comfortable and unobtrusive for the subjects in exam. In particular, the main goal is to develop real-time oriented algorithms able to minimize the errors and to obtain results as close as possible to those achieved by clinical instruments, especially with the aim of carrying out a constant home monitoring for the elderly. In this paper two contactless methodologies for the heart rate estimation are analyzed, exploiting different approaches. In the first we first apply face recognition methods on RGB videos, and we then use well known algorithms for the extraction of the heart rate, such as Eulerian Video Magnification, Independent Component Analysis, Principal Component Analysis and Skin Detection. In the second the cardiac frequency is extracted through a mmWave radar by applying two multiple-input multiple-output (MIMO) algorithms, one based on the Fast Fourier Transform (FFT) and one on the MUltiple SIgnal Classification (MUSIC) algorithm. For both approaches, the results are compared to those obtained using more standard instruments, such as a pulse oximeter, proving the accuracy and precision of the implemented systems.
Comparison of Video and Radar Contactless Heart Rate Measurements / Senigagliesi, L.; Ricciuti, M.; Ciattaglia, G.; De Santis, A.; Gambi, E.. - ELETTRONICO. - 1387:(2021), pp. 96-113. (Intervento presentato al convegno 6th International Conference on Information and Communication Technologies for Ageing Well and e-Health, ICT4AWE 2020 nel 2020) [10.1007/978-3-030-70807-8_6].
Comparison of Video and Radar Contactless Heart Rate Measurements
Senigagliesi L.;Ricciuti M.
;Ciattaglia G.;De Santis A.;Gambi E.
2021-01-01
Abstract
One of the most relevant aspects investigated by the scientific community in the healthcare field is the detection of vital parameters, such as the heart rate, through the use of contactless technologies, often considered more comfortable and unobtrusive for the subjects in exam. In particular, the main goal is to develop real-time oriented algorithms able to minimize the errors and to obtain results as close as possible to those achieved by clinical instruments, especially with the aim of carrying out a constant home monitoring for the elderly. In this paper two contactless methodologies for the heart rate estimation are analyzed, exploiting different approaches. In the first we first apply face recognition methods on RGB videos, and we then use well known algorithms for the extraction of the heart rate, such as Eulerian Video Magnification, Independent Component Analysis, Principal Component Analysis and Skin Detection. In the second the cardiac frequency is extracted through a mmWave radar by applying two multiple-input multiple-output (MIMO) algorithms, one based on the Fast Fourier Transform (FFT) and one on the MUltiple SIgnal Classification (MUSIC) algorithm. For both approaches, the results are compared to those obtained using more standard instruments, such as a pulse oximeter, proving the accuracy and precision of the implemented systems.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
