Among all the vital signs and physiological parameters, the respiratory rate (fR) is still considered the neglected vital sign. In the clinical scenario, occupational scenario and during sport and activities different methods can be used to record the respiratory waveform and thus the fR.In this paper, we investigated the performances of two different methods for the non-contact monitoring of respiratory waveform by using RGB video signal acquired from a single built-in high-definition webcam. Two different methods have been tested: the first based on the intensity change of video pixels and the second on the extraction of the optical flow from the video sequence.The results obtained so far from eight subjects in the time and frequency domain reveal that optical flow method is better performing than its counterpart. The maximum percentage error for the optical flow method in the frequency domain to calculate the breathing rate stands at less than 3% for all the subjects investigated by far. Likewise, in the time domain, the maximum mean absolute error (MAE) reported for the optical flow method stands at less than 1% in all the subjects.
Comparison of two methods for estimating respiratory waveforms from videos without contact
Massaroni C.;Schena E.;Silvestri S.;
2019-01-01
Abstract
Among all the vital signs and physiological parameters, the respiratory rate (fR) is still considered the neglected vital sign. In the clinical scenario, occupational scenario and during sport and activities different methods can be used to record the respiratory waveform and thus the fR.In this paper, we investigated the performances of two different methods for the non-contact monitoring of respiratory waveform by using RGB video signal acquired from a single built-in high-definition webcam. Two different methods have been tested: the first based on the intensity change of video pixels and the second on the extraction of the optical flow from the video sequence.The results obtained so far from eight subjects in the time and frequency domain reveal that optical flow method is better performing than its counterpart. The maximum percentage error for the optical flow method in the frequency domain to calculate the breathing rate stands at less than 3% for all the subjects investigated by far. Likewise, in the time domain, the maximum mean absolute error (MAE) reported for the optical flow method stands at less than 1% in all the subjects.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
