Seismic Prospecting Signal Denoise Disposal Technology Research

Along with development of the seismic survey work, the oil field exploration gradually changesdirection from the shallow portion to the depth portion, and from the flat land to the mountainousarea, the desert region. Because the gathering condition has been worse and worse, the noise containedin seismic data which gathered in seismic survey will increase, these noises and the information whichis related with the subterranean structure and lithology are mutually interwoven. Therefore, it is notsuitable to directly make the geologic interpretation using the field seismic data, needs to carry onthe digital processing to it, distills the useful information, thus provides the reliable material for theseismic survey geologic interpretation. Among them, the signal noise reduction is an especiallyimportant step in the digital processing, it is used to distill the useful information from the seismicdata, enhances the signal to noise ratio of the seismic data. This paper emphatically studied the noisereduction technology in the seismic survey signal, with which the wavelet transformation and theK-L transformation technology are combined.This article first briefly introduces the seismic survey's principle, production work, terminologyexplanation and signal noise, then the wavelet transform is elaborated in detail. And then the waveletthreshold value de-noising method and K-L transform de-noising method are studied, aiming attheir respective advantages and disadvantages, the improvements to each algorithm are made,proposing the translation invariant wavelet threshold value de-noising method and the space-timeplus obliquity adjustment processing algorithm based on the K-L transform. The experimentalresult indicated, utilizing these two improved algorithm to process the seismic data, the sectionplane noise has been removed well. Moreover, because signal and noise have the different spreadcharacteristic in each scale of the dyadic wavelet transform, and the construction signal can bepreferably obtained from the signal modulus maximum value using conjugate grads algorithm,this paper then use the dyadic wavelet transform modulus maximum value de-noising methodto process the model data and the seismic data. The experiments indicate that this de-noisingmethod can obtain the good effect, improving the seismic data quality.