Time and Frequency Domain Feature Extraction Method of Doppler Radar for Hand Gesture Based Human to Machine Interface

Abstract

In the development of hand gesture based Human to Machine Interface, Doppler response feature extraction method plays an important role in translating hand gesture of certain information. The Doppler response feature extraction method from hand gesture sign was proposed and designed by combining time and frequency domain analysis. The extraction of the Doppler response features at time domain is developed by using cross correlation, and the time domain feature is represented by using peak value of cross correlation result and its time shift. The Doppler response feature of frequency domain is extracted by employing a discriminator filter determined by the frequency spectrum observation of Doppler response. The proposed method was employed as a preprocessing for Continuous Wave (CW) radar output signals, which is able to relieve the pattern classification of Doppler response associated with each hand gesture. The simulation and laboratory experiment using HB 100 Doppler radar were performed to investigate the proposed method. The results show that the combination of all three features was capable of differentiating every type of hand gestures movement.

Hand Gestures Recognition Using Radar Sensors for Human-Computer-Interaction: A Review

“…CW radars can further be classified as Frequency-Modulated CW (FMCW) radars and single-frequency CW (SFCW) radars. Several studies also denote SFCW radars as Doppler radars [33] since their detection mechanism relies heavily on the Doppler phenomenon. SFCW radar sends a single-tone frequency signal and a shift in the frequency of the received signal occurs when it encounters the hand.…”

“…Four gestures were used, with an average accuracy of 94.5%. A study published by [33] presented a features extraction method, followed by cross-correlation and peak detector gesture classification. Another very recent study published by Yu and co-workers [57] reported 96.88 % accuracy for Doppler-radar-based HGR using spectrograms as input to the DCNN network.…”

A Survey about application based on RF signals

“…In the past decade, radar based HGR has witnessed a great climax and rapid development. From the type of radar, gesture recognition can be divided into ultra bandwidth pulse radar [1,2,3,4] and continuous wave radar [6,7,15,16,8]. For pulse radar, it mainly uses the original data-driven deep learning method to recognize gestures, and the related feature extraction work is less.…”

“…This paper mainly summarizes the RF applications based on radar. According to the working mode, the radar can be divided into pulse radar [1,2,3,4] and continuous wave radar [5,6,7,8]. Pulse radar intermittently transmits rectangular pulse periodic signals and receives reflected echo signals in the transmission gap to perceive the environment, but pulse radar has a blind area for short-range detection.…”

“…In the same year, a through wall human body contour capture system RF capture [11] was proposed to track the 3D position of a person's limbs and body parts, generate a coarse-grained human body contour, and distinguish different users through the contour information, which laid a foundation for subsequent human body posture estimation. Subsequently, with the rapid development of perception technology and deep learning based on RF signal, a series of applications such as sleep detection [12,13], fall detection [14], gesture recognition [6,7,1,2,3,4,15,16,8], radar imaging [17,5,11,18], physiological feature monitoring [12,19,20,21,22] , gait recognition [23,24] , track tracking [25,26] have been realized. It is widely used in video games, intelligent video surveillance, smart home, human-computer interaction, security and other fields, and has broad application prospects and economic value.…”