Seismic Horizon Tracking via Subspace Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing seismic full horizon tracking methods struggle with layer crossing phenomena in complex geological areas, failing to accurately interpret three-dimensional seismic data and resulting in time-consuming and inaccurate horizon tracking.
Innovation Solution
A method that involves acquiring three-dimensional seismic data, extracting horizon extreme points, dividing the sample space into sub-spaces, performing clustering using DBSCAN, establishing topological consistency, and fusing horizon fragments to achieve accurate full horizon tracking, avoiding layer crossing and improving tracking efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If existing seismic full horizon tracking methods are used, then the tracking process is simplified, but layer crossing phenomena occur in complex geological areas resulting in inaccurate tracking
Solution Approach 1:
The sample space is divided into multiple subspaces with overlapping portions, and clustering is performed independently in each subspace to obtain horizon fragments. This segmentation approach prevents layer crossing by localizing the clustering process while maintaining overall accuracy through the overlapping regions that provide continuity constraints.
2Measurement precision
If manual artificial tracking method is used, then the tracking can be performed step by step, but it is time-consuming and effort-consuming
Solution Approach 1:
The system performs automatic horizon tracking by extracting extreme points, dividing the sample space into subspaces, and applying clustering algorithms independently in each subspace. This self-service automation eliminates manual intervention while maintaining tracking quality through the structured subspace approach and overlapping region constraints.
3Extent of automation
If semi-automatic tracking method based on seed points is used, then some automation is achieved, but it still does not make full use of three-dimensional seismic data
Solution Approach 1:
The three-dimensional sample space is segmented into multiple subspaces with overlapping portions, allowing the clustering algorithm to process data locally while the overlaps ensure global consistency. This approach fully utilizes the three-dimensional structure of seismic data by treating each subspace as a independent clustering domain that contributes to the overall horizon tracking result.
Data Source
AI summary
There is disclosed in the present disclosure a seismic full horizon tracking method, a computer device and a computer-readable storage medium. The method includes: acquiring three-dimensional seismic data; extracting horizon extreme points from the three-dimensional seismic data to construct a sample space; equally dividing the sample space into a plurality of sub-spaces with overlapping portions, and performing a clustering process on the horizon extreme points in each sub-space to obtain horizon fragments corresponding to each horizon of the three-dimensional seismic data; establishing a topological consistency between the horizon fragments; and fusing the horizon fragments corresponding to each horizon of the three-dimensional seismic data based on the topological consistency, to obtain a full horizon tracking result of the three-dimensional seismic data. In the disclosure, a layer crossing phenomenon occurring in seismic full horizon tracking can be avoided, and a better full horizon tracking effect can be achieved.