Millimeter wave radar sensing gesture recognition method based on few sample learning

Abstract

A millimeter-wave radar sensing gesture recognition method based on few-sample learning comprises the steps of designing action gestures and radar parameters, building a millimeter-wave radar system platform, enabling a human body to stand in front of a platform collection position, carrying out gesture action change, transmitting a linear frequency modulation signal through a millimeter-wave radar, then receiving an echo signal containing gesture information, and carrying out gesture recognition. Performing frequency mixing processing on the transmitting signal and the receiving signal to obtain an intermediate frequency signal; then clutter suppression and 2D-FFT preprocessing are carried out on the intermediate frequency signals, distance-time and speed-time data sets of gesture actions are constructed, and a plurality of representative distance and speed features are selected from the distance-time and speed-time data sets to serve as a network input sample set; and optimizing SVM model parameters by using an improved grid search algorithm, selecting optimal parameters to realize data set training with few samples, and realizing gesture recognition. According to the invention, a high-accuracy gesture recognition effect is realized. The method is wide in application prospect and high in practicability.

Description

technical field [0001] The invention relates to the technical field of human-computer interaction, in particular to a millimeter-wave radar sensing gesture recognition method based on few-sample learning. Background technique [0002] With the rapid development of artificial intelligence technologies such as neural networks and deep learning, gesture action recognition technology has become one of the research hotspots in the fields of wireless perception, pattern recognition, computer vision, and signal processing. It has broad market prospects and far-reaching social application value. For example, in home life, people can use air gestures to remotely adjust the volume and switch programs without touching the device and pressing buttons, making the user experience more convenient and comfortable; in the field of car driving, drivers use designated dynamic gestures for driving operations. It can improve security; in terms of entertainment games, players can maximize innova...

AI Technical Summary

Problems solved by technology

[0005] At present, the gesture recognition method of millimeter wave radar is mainly based on deep learning algorithm, which has large data set dependence and computational complexity

Gesture recognition based on deep learning is a data-driven algorithm, which cannot effectively solve the problems of scarcity of training data and limited computing resources.

Generally speaking, a high-performance recognition network often needs to adopt a deeper network structure. The larger the parameter scale of the model, the more training samples required, the higher the computational complexity, and the greater the risk of over-fitting the model.

There are few public data sets for radar-based gesture recognition, most of which are based on self-test data or simulation data, and it is difficult to obtain a large amount of radar echo data within a limited time in the experiment

Images