Research On Gesture Recognition System Based On Millimeter Wave Radar

Traditional vision-based gesture recognition is susceptible to environmental factors such as light intensity,rain and snow which greatly reduces the accuracy of gesture recognition.In addition,vision-based data acquisition equipment is also vulnerable to attacks by criminals,and there is a risk of privacy leakage.In recent years,with the development of radar technology,it has been more and more widely used in military and civilian fields.This paper uses millimeter wave radar equipment to conduct research on gesture recognition.Feature information reflecting different gestures is extracted by analyzing the radar echo data of four gestures,and then this feature information is input to a classifier for classification and recognition.The paper mainly completed the following work and research:(1)The time-frequency analysis method is used to process the radar gesture echo data and then time-frequency power spectral density is used as the feature vector to input to the convolutional neural network for training and recognition.After experimental verification,this method can achieve a recognition accuracy of 86%.(2)The range doppler processing method is used to analyze the radar gesture echo data.Based on the obtained Range-Doppler map,the moving trajectory of the moving target corresponding to the maximum peak is extract as features which can achieve a accuracy of 92%.(3)The paper design a real-time gesture recognition system based on the client/server architecture which includes client data acquisition module,client data pre-processing module,server-side database,server-side multi-threaded concurrent solution,and communication protocol design for both parties to ensure system stability and scalability.