Micromotion Gesture Recognition Technology Based on Millimeter-wave Radar

作者：D.C. Cai, Yingchun Li, Zhaoqi Li, Fei Gao, Panpan Tang, Zheyi Li, Cong Wang, Jingchang Nan · 年份：2024 · DOI：10.1109/iceict61637.2024.10671018 · 研究领域：Hand Gesture Recognition Systems、Advanced Computing and Algorithms

Different from the traditional FMCW radar signal processing method, this paper mainly explores the application of the time-frequency analysis method of the reflected intermediate frequency of millimeter wave radar in fretting detection. The radar echo data are analyzed by FFT, wavelet transform and Hilbert yellow transform. This topic mainly uses human gesture targets as the micromotion feature to detect. Under the three designed gestures, the time-frequency analysis method proposed in this topic combined with a convolutional neural network achieves a high recognition success rate. The motion detection technology proposed in this paper can be applied to millimeter wave radar under THE FMCW system. It can detect and classify tiny motions under complex backgrounds and show good recognition ability under strong interference. It can be used in medical treatment, automobile and other fields to play the role of accurate identification.