Feature-Based Hand Gesture Recognition Using Two-Antenna Doppler Radar System

In this paper, we studied the hand gesture recognition algorithm based on features extraction under the single-frequency continuous wave and two-receiving antenna structure in mm wave radar. Through the radar parameters design and signal processing, the time-frequency spectrum can be generated by single frequency continuous wave and the phase distribution spectrum can be generated by the two-receiving antenna structure. 12 artificial design features are extracted from the time-frequency spectrum and phase distribution spectrum after OS-CFAR de-noising method. The support vector machine (SVM) is used to recognize the seven defined gestures from extracted features, and the recognition rate can reach 93.84%.