Self-similarity analysis on human backscattering in radar

The self-similarity of Human target backscattering in radar is analyzed. The human echo of typical radar signals (i.e., range profiles, micro-Doppler image and range-Doppler video sequence) is modeled by motion capture data. Then the self-similarity matrices (SSM) of the radar signals are constructed and compared. Since the SSM of range-Doppler video sequence exhibits periodic information in a more simple manner, it is further analyzed to demonstrate the angle-invariant characteristic of SSM. Finally, the periodicity of different activities (i.e., walking and running) are extracted from SSM and compared.