Abstract
Ultrasonic-based technologies could provide light-independent and high-resolution solutions in no-contact human–machine interaction. This paper proposed a gesture recognition system on the basis of ultrasonic frequency modulated continuous wave (FMCW) and ConvLSTM model. It employed a hardware layout of one transmitter and three receivers spatially installed at different directions. An FMCW signal emitted by the transmitter would be detected by the receivers after reflecting from the hand. Then, the range-Doppler maps (RDMs) of the received signal were obtained by processing 2D fast Fourier transform. The spatio-temporal features of the hand were extracted from RDMs by the ConvLSTM model, to facilitate the gesture classification. Experiments showed that a fine resolution of 0.005 m in distance and 0.03 m/s in velocity for hand movement, and an accuracy of 85.7% with a small size of 50 samples training for finger-movement gestures can be achieved, which verified the feasibility of the proposed system.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
