Multi-threshold Wavelet Packet-Based Method to Attenuate Noise from Seismic Signal

Abstract

Seismic data often contains different types of noise, and eliminating noise is one of the most important issues in seismic data processing. Noise on seismic data has influence on decline of SNR (signal to noise ratio) and resolution ratio of seismic data. Because of the reasons above, it is difficult to accurately obtain the geological structure by analyzing seismic signal. In view of limitation of selecting threshold value in the traditional wavelet analysis method, this paper proposes the multi-threshold wavelet packet method based on the frequency sequence. Through theoretical analysis and simulation of seismic signal, the de-noising processing is realized by using the multi-threshold wavelet packet de-noising method, the results show that the method can effectively filter out seismic noise and reserve the useful signal with the middle and high frequency. So the capability of noise reduction is superior to other traditional methods, which can effectively improve the resolution ratio of the seismic signal and establish the foundation for later inversion of geological structure.