Robust non-contacting human respiration signal detecting system by applying a new algorithm to minimize noise

In this paper, we conducted an experiment regarding a non-contacting respiration detection system by applying a new algorithm to minimize the noise effect. A Doppler radar in the 5.8GHz band with a series algorithm which includes a Savitzky-Golay Filter and a signal integration process to minimize the noise effect was used for the measurement of the human respiration signal at 1.2m distance. This signal processing algorithm, including the removal of DC signals from direct down-conversion and noise suppression, is faster than the previously reported system in the literature, and allows for a robust determination of the presence of the respiratory signals for a variety of situations. The results show the final signal magnitude ratio is at least 3.1 times bigger when the human being is present. compared to when the human being is absent. We have compared signals before and after the proposed signal processing algorithm and verified that the proposed system can be more robust to detect the human respiration signal even in a very noisy environment. We believe the proposed system can be very reliably applied to a detection of human being in many casualty incidents.