Through-Wall Detection of the Moving Paths and Vital Signs of Human Beings

Abstract

Detection of human activities in complex environments such as through wall by ultrawideband radar has many important applications in security, vital rescue, and so on. It is much more difficult to detect vital signs of moving human beings than static ones. In this letter, we build a model for moving targets and apply the time domain finite element method to simulate single-input multiple-outputs (SIMO) radar data. Human respiration is modeled by changing body size and physical parameters. The background removal is performed for radar data. Then, we use the back projection to reconstruct the consecutive target locations, which constitute the moving path, leading to a curve carrying vital signs in the radar image. Since SIMO radar data are multivariate, we use multivariate empirical mode decomposition (MEMD) and fast Fourier transform to separate and extract the respiratory characteristic frequencies. The reconstructed frequency coincides with that in the original model. The result shows that the combination of SIMO radar and MEMD can effectively identify the moving path of the human being behind the wall and extract vital signs.