Most existing non-contact monitoring systems are limited to detecting physiological signs from a single subject at a time. Still, another challenge facing these systems is that they are prone to noise artifacts resulting from motion of subjects, facial expressions, talking, skin tone and illumination variations. This study proposes an efficient non-contact system based on a digital camera to track the cardiorespiratory signal from a number of subjects (up to six persons) at same time with a new method for noise artefact removal. The proposed system relied on the physiological and physical effects as a result of the activity of the cardiovascular and respiratory systems, such as skin color changes and head motion. Since these effects are imperceptible to the human eye and highly affected by the noise variations, we used advanced signal and video processing techniques, including developing video magnification technique, complete ensemble empirical mode decomposition (EMD) with adaptive noise and canonical correlation analysis to extract the heart rate and respiratory rate from multiple subjects under the noise artifact assumptions. The experimental results of the proposed system had a significant correlation (PCC=0.9994, SCC=0.9987 and REME=0.32) when compared to the conventional contact methods (Pulse oximeter and Piezo respiratory belt), which makes the proposed system a promising candidate for novel applications.
Simultaneous Tracking of Cardiorespiratory Signals for Multiple Persons Using a Machine Vision System with Noise Artifact Removal https://www.embs.org/jtehm/wp-content/uploads/sites/17/2017/10/simultaneous.jpg 780 435 IEEE Journal of Translational Engineering in Health and Medicine (JTEHM) //www.embs.org/jtehm/wp-content/uploads/sites/17/2022/06/ieee-jtehm-logo2x.png