Technique for interactive wavelet extraction and decomposition in cepstrum domain

The method of wavelet extraction from a seismic record is a typical problem in the processing of seismic data. Provided that the seismic wavelet is known, many problems about seismic data processing and the inversion of seismic attributes can be solved completely. In order to solve this problem, many methods of wavelet extraction and decomposition have been introduced. But every method has its limitation and prerequisites. Based on systemic analysis of previous methods of wavelet extraction, this paper presents a method which uses the wavelet amplitude spectrum to decompose the wavelet into minimum and maximum phase components. The group of wavelets with the same amplitude spectrum and different phase spectrum can be found in the cepstrum domain. According to the criterion of maximum variance module and the previous information and processing objectives, with an interactive processing tool, users can select optimum wavelets among the wavelet group. Furthermore, the seismic resolution can be improved by using the extracted wavelet to do wavelet deconvolution and wavelet zero-phasing. Wavelets have been extracted from theoretical records and real seismic data, and the tests for improving resolution also have been carried out with the extracted wavelets. Examples showed the validity and effectiveness of this method.