Continuous Respiratory Rate Estimation Approach Based on Phase Features Using Array Radar in Home Sleeping Monitoring

Vital signs daily monitoring using a radar sensor is a popular and important contactless technique because of its advantage of penetrating the quilt, not leaking privacy, being sensitive to human micromotion, and so on. However, multiple people’s respiration rate estimation for the case of two people with arbitrary sleeping postures is still a challenging issue. In this article, we propose a continuous respiratory rate estimation approach based on phase features using array radar in home sleeping monitoring. The motivation of the proposed approach lies in that the phases can reflect the details of people’s breathing. Specifically, we acquire the possible clusters of people and the corresponding features of these clusters to determine the people number and their locations. Then, we obtain the possible phases of people and select candidate respiratory phases from these possible phases according to the phase shape similarity feature and the respiratory intensity feature. Subsequently, we calculate the candidate respiratory phases’ rates and associate people’s respiratory rates based on the respiratory frequency matching strategy which can reduce the mutual interference among multiple people. The experimental results show that the proposed approach can achieve an average mean accuracy of 97.4% in one-person cases, and 96.94% in two-person cases and show robustness to different sleeping positions, people number changing, and people sleeping postures.