A Fast Dynamic Gesture Recognition Method Based on 3D Trajectory

With the arrival of the 5G era, gesture recognition technology plays an important role in the smart home system. This paper adopts a Frequency Modulated Continuous Wave (FMCW) radar to design a fast gesture recognition system. Specifically, this paper proposes a dynamic gesture recognition method based on 3D trajectory, which is generated by fusing the range, angle, and time information, and corrected by class targets suppression through Doppler. Then the convolutional neural network (CNN) recognizes and classifies the dynamic gesture with the 3D trajectory. Finally, experiments are carried out to verify the effectiveness of the proposed method.