움직이는 사람에 대한 비접촉식 생체신호추정 연구

Abstract

The non-contact vital sign estimation of an individual using radar sensors has been widely researched. However, in the case of an individual with undesirable rigid-body motion of the torso, non-contact vital sign estimation may fail owing to extreme phase fluctuations caused by the larger rigid-body motion of the torso than the micro motion of the vital sign. To address this problem, we propose a rigid-body motion compensation method using the complex echo signals of the IR-UWB radar for the vital sign estimation of an individual with rigid-body motion. The proposed method extracts a complex signal vector using principal component analysis from complex echo signals in a radial range and pulse time domain. Next, compensation for the rigid-body motion is implemented by minimizing the entropy of the spectrum for the complex signal vector. In simulations using a point-scatterer model and experiments using IR-UWB commercial radar, we verified that our proposed method is capable of performing successful vital sign estimation of an individual with undesirable rigid-body motion.