Choosing the Analyzing Wavelet for Seismic Data Filtration

The mathematical apparatus of wavelet transform includes wide collection of base wavelet functions and can be used for nonstationary signal processing including wave fields of elastic vibrations spreaded in geological medium. The possibility of localization of anomalies of signal component in seismic fields is important characteristic of wavelet transform on conditions that analyzing wavelet matches with the wave impulse of elastic vibrations. The results of wavelet analysis of seismic data after reverse transition from time-scale space of wavelet coefficients to space of time and coordinates are given in this article. The seismic data processing is carried out using the discrete wavelet transform based on Mallat pyramidal algorithm. The efficiency of wavelet filtration with decomposition of signal on detailed layers or wavelet coefficients muting by significant level with use of different classes of analyzing wavelets is shown on real seismic data. Wavelet filtration based on detailing layers of different levels with use of smooth Daubechies, Coiflet, and Symlet wavelets allows marking out and increasing the target signal component of seismic field, and muting the low-frequency regular noise (surface waves) and random noise with no significant change of local features of signal component in time/coordinate space.