A novel time-frequency analysis method based on HHT for finer-grained human activity using SFCW radar

Recently the increasing anti-terrorism and disaster rescue operations underscore the requirement for effective detection and recognition of finer-grained human activity. Currently, the most effective technology is usually realized by exploiting time-frequency analysis to extract the micro-Doppler (MD) signature difference formed by different parts of the body movement. However, for inherently nonlinear and non-stationary human motion signal, the traditional signal processing technique that assume linearity and stationary such as STFT are not optimal. In this paper, we propose a novel comprehensive range accumulation time-frequency transform (CRATFR) based on the Hilbert-Huang Transform (HHT) to obtain micro-Doppler signatures of human finer-grained human activity signal detected by ultra-wide band (UWB) radar for motion analysis and recognition. Therefore, this paper firstly adopts the SFCW radar to conduct the detection of finer-grained human activity. Then, according to the SFCW radar signal characteristics of human motion, the CRATFR was adopted to obtain the MD signature. Finally, the experiments at 3m and 10m position were conducted to demonstrate the validity of the proposed method.