Signal Processing Based on Gesture Recognition System

Abstract

In the process of using gestures to unlock a mobile phone, a variety of algorithms can be used for data matching, such as Dynamic Time Warping algorithm (DTW) and Support Vector Machine algorithm (SVM). The processing method of the gesture signal will greatly affect the accuracy of gesture authentication before the algorithm is used for matching. Therefore, this article proposes a more general data signal processing method. This method mainly uses threshold-based data separation method, FIR filter to remove noise from data and data fitting training for processing. The result shows that the use of this signal processing method can make the gesture recognition rate higher than traditional algorithms, that is, a better recognition state can be achieved. It is expected to lay the foundation for the development of more accurate unlocking methods if this method is applied to the signal processing of mobile phone unlocking.