Simulation of human microDopplers using computer animation data

The capability to detect, track and monitor human activities in non-line-of-sight environments is an important component of security and surveillance operations. The movement of the human body and limbs result in unique microDoppler features which can be exploited for identification and classification of different types of human motions. In this paper, a method to simulate the microDoppler signatures of complex human motions from virtual reality animation data is presented. The simulation method is tested on three types of human motions - walking, running and crawling. The simulated Doppler spectrograms are compared with measurement results. Additional simulation data are presented to show patterns over a long time duration and the effect of frequency and viewing angle.