Seismic Data Residual Migration via Parsimonious Image Decomposition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current seismic data processing techniques for fast migration often discard beams, leading to inefficiencies in representing energy and estimating dips, which limits the accuracy of depth migration and image formation in seismic exploration.
Innovation Solution
A software-implemented method for seismic data processing that involves initial migration, decomposition of data into beams using a matching pursuit scheme to select dominant dips, and residual migration to form a parsimonious image representation, allowing for efficient and accurate dip estimation and image formation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional fast migration discards beams to rapidly image important features, then processing speed is improved, but manufacturing precision of the seismic image deteriorates
Solution Approach 1:
The patent transforms the seismic data from time-domain to frequency-domain parameters, enabling decomposition into frequency-wavenumber components. This parameter transformation allows selective processing of different frequency bands and wavenumber ranges, achieving both computational efficiency and imaging accuracy by processing only the most significant frequency-wavenumber components rather than discarding beams arbitrarily
Solution Approach 2:
The patent segments the seismic data into multiple frequency bands and wavenumber ranges, creating a decomposed representation where each segment can be processed independently. This segmentation allows the migration algorithm to focus computational resources on dominant frequency-wavenumber components while maintaining the ability to reconstruct the complete image with high accuracy
2Device complexity
If a limited number of fixed dips are used in migration, then device complexity is reduced, but measurement precision of dip estimation deteriorates
Solution Approach 1:
The patent replaces fixed, static dip angles with dynamic dip estimation that adapts to the local characteristics of the seismic data. By performing dip estimation in the frequency-wavenumber domain and using the relationship between wavenumber and dip angle, the algorithm dynamically determines the appropriate dip for each frequency-wavenumber component, achieving high accuracy without requiring a large number of fixed dip samples
Solution Approach 2:
The patent moves the dip estimation problem from the spatial domain to the frequency-wavenumber domain, adding a dimensional transformation that simplifies the complexity. In this transformed domain, dip information is directly encoded in the wavenumber components, allowing accurate dip estimation through simple spectral analysis rather than complex spatial sampling of multiple fixed dips
3Loss of time
If beams are discarded in fast migration, then loss of time is reduced, but loss of information about seismic energy increases
Solution Approach 1:
The patent changes from processing in the time-space domain to the frequency-wavenumber domain, where energy distribution can be selectively analyzed and processed. This parameter transformation enables the algorithm to identify and process only the significant frequency-wavenumber components that contain the majority of the seismic energy, maintaining energy representation accuracy while reducing processing time
Solution Approach 2:
The patent extracts and processes only the dominant frequency-wavenumber components that contain the significant seismic energy, rather than processing or storing all beam data. By extracting the essential energy-carrying components in the frequency-wavenumber domain, the algorithm maintains accurate energy representation while discarding redundant information, achieving both time efficiency and information preservation
Data Source
AI summary
A technique for performing a fast residual migration of seismic data through parsimonious image decomposition is presented. In one aspect, the technique includes a software-implemented method for processing a set of seismic data includes through parsimonious image decomposition. Other aspects of the technique include a program storage medium encoded with instructions that, when executed by a processor, perform such a method or a computing apparatus programmed to perform such a method.