A new method of dynamic gesture recognition using Wi-Fi signals based on Adaboost

摘要

In order to find a method for dynamic gesture recognition without adjusting parameters according to different gestures and different training environment, a new method based on Adaptive Boosting (Adaboost) classification method is proposed to implement dynamic gesture recognition in this paper. The combination of Gaussian filter and Median filter is applied to preprocess the data. Six predefined dynamic gestures were tested in our experiment. A large number of experiments show that our method can achieve high accuracy of gesture recognition, with the average recognition rate of 95.20%. Also, the comparison between the proposed and the traditional classification method was discussed. According to the obtained results, the method presented in this paper is more effective with less time cost for dynamic gesture recognition.