A Novel Recognition Algorithm of Human Micro-motion by Using Continuous Wave Radar

Human micro-motion can affect the detection of cardiopulmonary signals. To recognize the interference of human micro-motions, a 24 GHz continuous wave (CW) radar is utilized to collect signals of different micro-motions. The recognition is performed based on time-frequency features extracted by a short-time iterative adaptive approach (STIAA). STIAA can obtain a higher resolution compared to conventional short-time Fourier transform (STFT). Based on the time-frequency features, the transfer learning technique with pre-trained convolutional neural networks is applied to recognize five human micro-motions. The experiment results demonstrate that the performance of the proposed scheme is promising.