Enhanced Non-Contact Vital Sign Monitoring via Depth Camera-assisted Millimeter-Wave Radar

Non-contact vital sign monitoring with mmWave radar has been regarded as an effective method for identifying lung and heart diseases due to its strong penetration and non-invasive capabilities. Nevertheless, relying solely on millimeter-wave radar for vital sign monitoring is susceptible to interference from multipath reflections, which degrades the measurement accuracy. To address this issue and enhance measurement accuracy, we propose a system design that leverages the depth camera for precise target location and subsequently employs dynamic beamforming techniques to effectively monitor the specified region. The experimental results demonstrate that our proposed system achieves an average error of 2.44 beats per minute (BPM) for heart rate monitoring and 0.22 breaths per minute (BPM) for respiratory rate monitoring. Compared to systems without beamforming, our system reduces the mean error for heart rate monitoring by 4.21 BPM and for respiratory rate monitoring by 0.51 BPM.