3D Seismic Fault Detection via Vector Correlation Difference
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for detecting faults in seismic images fail to accurately identify subtle variations, relying on 2D approaches that ignore 3D information and provide coarse results, making it difficult to detect subtle faults and improve fault network and geobody definition.
Innovation Solution
A computer-implemented method processes 3D seismic images by determining maximal correlations between vectors of seismic values associated with points on either side of a plane, shifting these vectors to calculate a difference value that aids in fault detection, providing an easier interpretation for human eyes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If classical edge detection methods or conventional 2D approaches are used to detect faults, then the processing is simple and fast, but the detection precision is low and subtle faults are missed
Solution Approach 1:
The patent transitions from 2D fault detection to 3D fault detection by analyzing seismic images in three dimensions. The method computes fault indicators using 3D neighborhoods and tilted planes at multiple dip and azimuth orientations, enabling detection of subtle faults that 2D methods miss while maintaining computational feasibility through systematic 3D analysis.
Solution Approach 2:
The patent segments the 3D seismic image into local neighborhoods around each center point. For each neighborhood, it divides the space into connected sets on opposite sides of tilted planes, allowing localized correlation analysis that captures subtle fault signatures without requiring complex global processing.
2Measurement precision
If 3D information is fully utilized to detect subtle faults, then the detection precision improves, but the computational complexity increases
Solution Approach 1:
The patent performs correlation analysis on selected connected sets of points rather than all possible point combinations. By focusing computation on opposite sides of tilted planes where faults are most likely to manifest, it achieves adequate fault detection without exhaustive 3D analysis, balancing precision and computational power.
3Measurement precision
If the fault detection relies on maximum or minimum values of computed results, then the processing is simple, but the ability to detect subtle variations is poor
Solution Approach 1:
The patent enhances subtle fault variations by computing difference values between correlations on opposite sides of tilted planes. These difference values create enhanced contrast in the output, making subtle faults visible through color or intensity variations that are much more detectable by the human eye than the original subtle seismic variations.
Data Source
Figure 1
Figure 2a
Figure 2b
AI summary
The present invention relates to a method to process a three dimensional seismic image. The method comprises, for a center point and for a plane comprising said center point, said plane defining a first connected set of points, a second connected set of points, determining a first maximal correlation between a first vector of the seismic values associated with at least a point of the seismic image having a distance to the center point below the predetermined distance and on a first line, said plane comprising said first line, and a second vector of the seismic values associated with points on a second line parallel to the first line, the second vector being shifted by a plurality of shifting distances; determining a second maximal correlation between the first vector, and a third vector of the seismic values associated with points on a third line parallel to the first line, the third vector being shifted by a plurality of shifting distances; and /e/ determining a difference value based on at least a difference between the first maximal correlation and the second maximal correlation.