基于奇异值分解的倾角扫描叠加

Dip-scanning Stacking Method on Basis of Singular Value Decomposition

Singular value decomposition (SVL)) is one of the new approaches in seismic data processing, nevertheless. it can't enhance non-horizontal events effectively; and thus there is introduced in this paper a dip-scanning stacking method on the basis of SVD. A group of (dip-scanning) parameters for linear time shift is selected according to seismic data for each dip. Then the data are filtered using SVD method, and inverse linear time shift is applied to restore the data to their original coordinates Finally. the results from all dips are added to form the final output data. The method can enhance signal-to-noise ratio whether there are horizontal or non-horizontal events in the data. The computed result of the synthetic data demonstrates the feasibility and effectiveness of the method. A discussion is presented also on some other relative problems in actual data processing.