Research On Resolution Enhancement Method With Adaptive Lateral Constraint Based On Denoising Filter

The seismic data with high signal-to-noise ratio(SNR)and high resolution have become the basic data of high precision seismic exploration in recent years.One of the most important factors that affect the SNR of seismic data is random noise.Prediction filtering is the most classic random noise reduction method.However,there are two important problems of conventional prediction filtering: one is that the assumptions of noise model are inconsistent before and after filtering;the other is poor ability of signal-preserving.To overcome the problems of prediction filtering,we propose an inversion-based t-x domain random noise reduction method in this paper.Besides,the method is further extended to a 3D t-x-y version.Synthetic data and field data examples demonstrate that our proposed method has equal ability of noise attenuation with prediction filtering method,but performs better in signal-preserving than prediction filtering method.Deconvolution is an important method to enhance the resolution of seismic data.Traditional linear deconvolution technique cannot greatly improve the resolution of seismic data and it has been replaced by nonlinear sparse deconvolution method.However,conventional nonlinear deconvolution method is a trace-by-trace deconvolution method,which can easily harm the lateral continuity of seismic events.Besides,when there is noise in the seismic data,the method can be very unstable.To overcome the problem of conventional nonlinear deconvolution method,we use the prediction error filter which can be solved in the new random noise attenuation procedure as a lateral constraint and add it into the objective function of sparse deconvolution.Then we can get the multi-trace sparse deconvolution method.The new method cannot only preserve the lateral continuity of the deconvolution result effectively but also can keep the SNR of seismic data.