A Dynamic Continuous Hand Gesture Detection and Recognition Method with FMCW Radar

Abstract

In this paper, a continuous dynamic hand gesture detection and recognition method is proposed using a frequency modulated continuous wave (FMCW) radar. Specifically, we collect the raw radar data to estimate the radar intermediate frequency (IF) signal, and construct the range-time map (RTM) and Doppler-time map (DTM) with 2-Dimensional Fast Fourier Transform (2D-FFT). Then, we propose a hand gesture detection method, which obtains the amplitude of the normalized hand gesture and uses a threshold to effectively segment the continuous hand gesture. Finally, the hand gesture is recognized by the proposed Fusion Dynamic Time Warping (FDTW) algorithm based on the central time-frequency trajectory. Experiments with radar data show that the accuracy of the proposed hand gesture detection method can reach 96.17%, and compared with the traditional recognition algorithm, the proposed recognition algorithm can significantly improve the recognition accuracy rate (hand gesture average recognition accuracy rate can reach 94.50 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">%</sup> ) with the time complexity reduced by more than 50%.