Wi-Fi环境下基于改进EMD降噪的手势识别

he gesture recognition scheme based on Wi-Fi signal does not require the user to wear the device, and there is no requirement for the illumination intensity and noise decibel in the environment. Therefore, this gesture recognition technology has become a research hotspot in recent years. However, there are multiple electromagnetic signals in the frequency band in which Wi-Fi signals work. These signals will interfere the gesture recognition in the Wi-Fi environment and make the collected gesture data contain noise. In order to reduce the influence of noise on gesture recognition, this paper combines the noise reduction rules that can keep the signal continuity with the custom threshold function, which makes the signal smoother after the EMD noise reduction. Then, in order to avoid the difficulty of classification caused by the excessively high data dimension of gesture, principal component analysis (PCA) algorithm was used to extract the principal component to reduce the data dimension. Finally, the reduced dimension data is sent to support vector machine for training and testing. In the untrained position, the recognition rate of the gesture recognition system proposed in this paper can reach 85.88%.