Methodology of Wavelet Processing of Seismic Data with Decomposition on Detailed Layers

Methodology of seismic data processing and analysis with use of wavelet filtration based on discrete wavelet decomposition of signal on detailed layers is considered. Discrete wavelet transform is implemented using the Mallat pyramidal algorythm. Level n detailed layer is formed as a signal reconstructed from level n detailed coefficients while wavelet coefficients of all other levels are zero. As all transformations are linear, complex wave field is decomposed on layers with different energy and frequency characteristics. The first detailed layer contains high-frequency components of initial data; the next layers contain more smooth (low-frequency) components. It is allowed and possible to construct additional detailed layers of wavelet decomposition if not only smooth wavelet coefficients but also detailed coefficients on each level are subjected to the discrete wavelet transform. It is important that wavelet filtration based on wavelet decomposition detailed layers makes it possible to keep local anomalies of signal component. Efficiency of methodology of wavelet decomposition on detailed layers under dynamic processing is demonstrated on real seismic data.