Breathing Rate Estimation Behind Walls Exploiting a UWB Radar Sensor

Ultra Wideband (UWB) radar sensors are reliable tools deployed by First Responders (FRs) during search and rescue operations to detect the signs of life of trapped victims under debris or behind walls. In this paper, we investigate the problem of vital signs estimation, specifically Breathing Rate (BR) behind walls, thus proposing a tool that can be used by FRs for situation awareness and victim prioritization. The knowledge of the BR in a search and rescue operation is of paramount importance as it can reveal insights into the trapped victim's health condition. This work builds upon an openly accessible dataset comprising seven lying down and nine standing humans located between 1 and 4 m away from the radar sensor. This dataset has been proposed to accelerate the research work in the field, as it includes labeled data denoting i) human presence and absence, ii) the distance between the sensor and the victim, as well as iii) the victim's BR. Herein, the Fast Fourier Transform (FFT) is employed to transform the signal to the frequency domain, estimate the BR and benchmark it against the ground truth measurements. The results are a first validation of the proposed approach, as without any signal processing, the BR is estimated with a mismatch lower than three breaths per minute in 73% of the cases.