Iterative dip-steering median filter for seismic data processing

Patent metadata

Field	Value
Publication number	US-9429668-B2
Application number	US-201213714654-A
Country	US
Kind code	B2
Filing date	Dec 14, 2012
Priority date	Dec 15, 2011
Publication date	Aug 30, 2016
Grant date	Aug 30, 2016

Abstract

An iterative dip-steering median filter is provided for random noise attenuation in seismic data where conflicting dips are indicated in the data. A number of dominant dips inside a processing window or sample of the data are identified by a Fourier-radial transform in the frequency-wavenumber domain. A median filter is then applied along the dominant dip to remove noise, and the remaining signal after filtering is retained for further median filter iterations. Iterations are repeated to apply the median filter along the most dominant dip in the remaining data. The processing continues in subsequent iterations until all selected dips have been processed. The remaining signal of each iteration is then summed for final output.

First claim

What is claimed is:

1. A computer implemented method of processing a pre-stack seismic gather, in which time-spatial processing results of the pre-stack seismic gather indicate conflicting dips in subsurface formation structure, to attenuate random noise in the pre-stack seismic gather, comprising the computer implemented processing steps of:
(a) assembling the time-spatial processing results of the pre-stack seismic gather in the computer to form a series of overlapping time-spatial sample windows of the pre-stack seismic gather;
(b) transforming an individual one of the time-spatial sample windows of the pre-stack seismic gather into a frequency-wavenumber domain;
(c) transforming the individual time-spatial sample window from the frequency-wave number domain into a Fourier-radial domain;
(d) determining a map of peak values of the individual time-spatial sample window in the Fourier-radial domain;
(e) transforming selected ones of the peak values of the individual time-spatial sample window in the Fourier-radial domain to indicate dominant dips in subsurface formation structure in the individual time-spatial sample window;
(f) applying a median filter to the individual time-spatial window along a selected one of the indicated dominant dips in subsurface formation structure to attenuate noise from the individual time-spatial sample window; and
(g) storing the remaining signal in the median filtered individual time-spatial sample window;
(h) determining whether each of the indicated dominant dips in subsurface formation structure for the individual time spatial sample window has been selected; and
(i) if not, selecting another indicated dominant dip in subsurface structure and repeating steps (f) and (g) along the selected another indicated dominant dip in subsurface formation structure; or
(j) if so, forming a sum of the stored remaining signal of the median filtered individual time-spatial sample window along the selected dominant dips in subsurface formation structure;
(k) determining whether each of the sample windows of the assembled pre-stack seismic gather have been selected; and
(l) if not, selecting an additional one of the sample windows of the assembled pre-stack seismic gather and returning to step (b) for performing steps (b) through (h) on the data of the additional one of the assembled pre-stack seismic gather; or
(m) is so, storing formed sums as a noise attenuated median filtered pre-stack seismic gather for each of the overlapping time-spatial sample windows; and
(n) displaying the noise attenuated median filtered pre-stack seismic gather each of the overlapping sample windows for more accurate representations of the seismic survey results in areas of interest.

2. The computer implemented method of claim 1 ; wherein the step of determining a map of peak values comprises the step of:
determining a sum for each of the samples in the Fourier-radial domain.

3. The computer implemented method of claim 2 , wherein the step of applying a median filter comprises the step of:
applying a median filter to the dips in subsurface formation structure in a descending sequence of the dips.

4. A data processing system for processing a pre-stack seismic gather, in which time-spatial processing results of the pre-stack seismic gather indicate conflicting dips in subsurface formation structure, to attenuate random noise in the pre-stack seismic gather, the data processing system comprising:
a data storage memory storing the time-spatial processing results of the pre-stack seismic gather;
a processor for performing the steps of:
(a) assembling the time-spatial processing results of the pre-stack seismic gather in the computer to form a series of overlapping time-spatial sample windows of the pre-stack seismic gather;
(b) transforming an individual one of the time-spatial sample windows of the pre-stack seismic gather into a frequency-wavenumber domain;
(c) transforming the individual; time-spatial sample window from the frequency-wavenumber domain into a Fourier-radial domain;
(d) determining a map of peak values of the individual time-spatial sample window in the Fourier-radial domain;
(e) transforming selected ones of the peak values of the individual time-spatial sample window in the Fourier-radial domain to indicate dominant dips in subsurface formation structure in the individual time-spatial sample window;
(f) applying a median filter to the individual time-spatial sample window along the indicated dominant dips in subsurface formation structure to attenuate noise from the individual time-spatial sample window; and
(g) storing the remaining signal of the median filtered the individual time-spatial sample window;
(h) determining whether each of the indicated dominant clips in subsurface formation structure for the individual time-spatial sample window has been selected; and
(i) if not, selecting another indicated dominant dip in subsurface structure and repeating steps (f) and (g) along the selected another indicated dominant dip in subsurface formation structure; or
(j) if so, forming a sum of the stored remaining signal of the median filtered individual time-spatial sample window along the selected dominant dips in subsurface formation structure:
(k) determining whether each of the sample windows of the assembled pre-stack seismic gather have been selected; and
(l) if not, selecting an additional one of the sample windows of the assembled pre-stack seismic gather and returning to step (b) for performing steps (b) through (h) on the data of the additional one of the assembled pre-stack seismic gather; or
(m) if sp, storing the formed sums as a noise attenuated median filtered pre-stack seismic gather for each of the overlapping time-spatial sample windows; and
a display which forms an output record of the noise attenuated median filtered pre-stack seismic gather for each of the overlapping windows for more accurate representations of the seismic survey results in areas of interest.

5. The data processing system of claim 4 , wherein the processor in performing the step of determining a map of peak values performs the step of:
determining a sum for each of the samples in the Fourier-radial domain.

6. The data processing system of claim 4 , wherein the processor in performing the step of applying a median filter performs the step of:
applying a median filter to the dips in subsurface formation structure in a descending sequence of the dips.

7. A computer implemented method of processing a post-stack seismic section, in which time-spatial processing results of the post-stack seismic section indicate conflicting dips in subsurface formation structure, to attenuate random noise in the post-stack seismic section, comprising the computer implemented processing steps of:
(a) assembling the time-spatial processing results of post-stack seismic section in the computer to form a series of overlapping time-spatial sample windows of the post-stack seismic section;
(b) transforming an individual one of the time-spatial sample windows of the post-stack seismic section into a frequency-wavenumber domain;
(c) transforming the individual time-spatial sample window from the frequency-wavenumber domain into a Fourier-radial domain;
(d) determining a map of peak values of the individual time-spatial sample window in the Fourier-radial domain;
(e) transforming selected ones of the peak values of the individual time-spatial sample window in the Fourier-radial domain to indicate dominant dips in subsurface formation structure in the individual time-spatial sample window;
(f) applying a median filter to the individual time-spatial sample window along a selected one of the indicated dominant dips in subsurface formation Structure to attenuate noise fro

Assignees

• Saudi Arabian Oil Co

Inventors

• Huo Shoudong
• Zhu Weihong

Classifications

• G01V1/364 Primary

  Seismic filtering (G01V1/37 takes precedence) · CPC title

• G01V2210/3248

  Incoherent noise, e.g. white noise · CPC title

• G01V1/301 Primary

  for determining seismic cross-sections or geostructures · CPC title