Multi-Target Human Breathing Detection System

In this paper, a multi-target human breathing detection system is proposed. For the problem of azimuth estimation of multi-targets, the multi-signal classification (MUSIC) algorithm and the General Liner Combination Based Robust Minimum Variance Distortionless Response (GLCMVDR) algorithm are combined, and the combined algorithm can accurately estimate the azimuths of multiple human targets. For the breathing signals of multiple human targets superimposed on each other in space, the feature matrix Joint Approximate Diagonalization of Eigenmatrix (JADE) blind source separation algorithm is used for breathing detection of multiple targets, which can separate the breathing signals of each target from the superimposed signals. The effectiveness of the multi-target human breathing detection system in this paper is verified by multi-target detection experiments.