Micro-motion gesture recognition method based on millimeter-wave radar and convolutional neural network

Abstract

The invention belongs to the technical field of human-computer interaction, and particularly relates to a micro-motion gesture recognition method based on millimeter-wave radar and a convolutional neural network. The method mainly comprises the following steps: designing radar parameters and micro-motion gestures according to an application scene; using a millimeter-wave radar for periodically transmitting linear frequency modulation signals with determined radar parameters and receiving echo signals reflected by hands of a human body, and carrying out ADC sampling after difference frequency is carried out on the echo signals and transmitted signals to obtain digital intermediate frequency signals; processing the digital intermediate frequency signal, and calculating characteristic parameters of the micro-motion gesture; selecting a certain feature, and establishing a data set of multiple gestures; designing a convolutional neural network for the millimeter-wave radar feature image, and inputting a gesture data set for training to obtain a classification model; and calling the classification model to realize classification and recognition of various gestures. The method is high inpracticability, can be applied to the fields of smart home, air input, sign language translation, mechanical control, VR, AR and the like, and is wide in application prospect.

Description

[0001] Invention field [0002] The invention belongs to the technical field of human-computer interaction, and specifically relates to a micro-motion gesture recognition method based on millimeter wave radar and convolutional neural network. Background technique [0003] With the rapid development of the Internet of Things and intelligent devices, the way of human-computer interaction is constantly changing, from the early key-pressing method to the current touch screen, voice interaction and non-contact action interaction methods. As a non-contact human-computer interaction method, gesture recognition has very important application value in smart home, air input, sign language translation, mechanical control, VR, AR and other fields. However, the existing gesture recognition methods have some major problems: the gesture recognition methods based on visible light, infrared and other image information have high power consumption, high computational overhead, low efficiency of featu...

AI Technical Summary

Problems solved by technology

However, there are some major problems in existing gesture recognition methods: gesture recognition methods based on visible light, infrared and other image information have high power consumption, high computational overhead, low efficiency of feature extraction, limited image processing model capabilities, and are susceptible to light and occlusion. factors, and there is a risk of privacy leaks; the ultrasonic-based gesture recognition method has a small beam angle, low resolution, is susceptible to interference and background environments, is difficult in signal processing, high in computing costs, and has large differences in product specifications; based on Wi-Fi The gesture recognition method of radar and radar has low resolution and less information

Existing gesture recognition methods basically do not have the ability to recognize micro-movement gestures within a few centimeters

Images