A novel measurement framework of vital signs in cluttered environments using 4D mmWave radar point clouds

• Published in: Measurement science & technology Vol. 36; no. 4; p. 45109
• Main Authors: Tu, Silong, Liu, Zhenyu
• Format: Journal Article
• Language: English
• Published: 30.04.2025
• ISSN: 0957-0233; 1361-6501
• DOI: 10.1088/1361-6501/adc3b4

Non-contact measurements of human vital signs using millimeter-wave radar have a broad application prospect in many scenarios. However, accurately measuring vital signs in cluttered environments remains a challenge due to the presence of multiple interferences, which affect subject localization and submerge subtle vital signals. Inspired by the high resolution of millimeter-wave radar point clouds, a vital signs measurement framework based on 4D radar is proposed. This framework can distinguish human-related point clouds and obtain precise vital signs in cluttered environments. First, considering the characteristics of periodic changes in vital signs, a coarse-to-fine point cloud selection strategy is introduced to localize the subject and extract the phase signal. The human-related point clouds are distinguished by autocorrelation analysis method, followed by a phase signal screening method to select phase signal with strong vital signs. Second, considering the order of mode extraction and the differences in displacement changes caused by heartbeat and respiration, a fast successive variational mode decomposition method is proposed. This method utilizes the center frequency of the mode as the convergence criterion to efficiently separate respiratory and heartbeat motion patterns. Experiments conducted in a cluttered indoor environment demonstrate that the proposed framework effectively mitigates the effects of severe clutter interference, achieving respiratory and heartbeat rate accuracies of 98.44% and 98.14%, respectively.