Research on Time-frequency Feature Fusion Method based on Dynamic Gesture Identity Authentication

Abstract

In this paper, three kinds of time-frequency feature fusion methods based on dynamic gesture sensor are analyzed. First, the time-domain feature of sensor data is extracted; second, the cepstrum feature, the power spectrum feature of Fourier Transform (FT) and the wavelet transform (WT) feature of gesture data are extracted and trained using SVM classifier model. The influence of feature extraction methods based on the fusion of time-domain features and three frequency-domain features on dynamic gesture authentication is compared, the experimental results show that the fusion of time domain features and frequency domain features of wavelet transform with fewer feature points has the best effect, which improves the accuracy of gesture authentication, So the method is worth promoting.