Fiber Optic Vibration Signal Feature Extraction and Recognition Based on Wavelet Transform

This paper conducted a research on the feature extraction of the fiber optic vibration signal and vibration signal identification based on optical fiber security warning system.It used the multiresolution of wavelet from the perspective of wavelet multi-scale decomposition to extract the details of vibration signal according to the relationship between time-frequency characteristics of wavelet transform and hierarchical decomposition,so that the features of vibration signals of different frequency band energy distribution can be reflected,on the basis of which,support vector machine was used to classify the vibration signals.The test result shows that compared with the existing methods,this method has the advantage of high accuracy.It provides a practical method with high performance for the identification of the fiber optic vibration signal.