3D Norm-Zero Optimization for Geophysical Data Smoothing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Geophysical data processing techniques, such as seismic inversion, face challenges in accurately modeling subsurface structures due to noise and complexity, particularly in blocky layered structures, where traditional filters like median filters may not effectively smooth data while preserving edges.
Innovation Solution
The implementation of a 3D norm zero optimization process, which groups raw geophysical data into subsets and applies a 3D norm zero objective energy function to iteratively smooth the data, removing noise while preserving edges, using a two-step iterative process that computes auxiliary values and updates input data with a beta value increase until convergence.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If traditional median filters are used to smooth geophysical data, then noise is reduced, but edges and important geological features are blurred or lost
Solution Approach 1:
The patent applies different treatment to different regions of the data by using adaptive filtering that adjusts local window sizes based on geological feature detection. The filter becomes more aggressive in homogeneous regions (removing noise) while preserving detail in regions with high gradient variations (edges and features), thus achieving local optimization of both noise reduction and edge preservation
Solution Approach 2:
The patent dynamically changes filtering parameters (window size, threshold values) based on local data characteristics. By adapting these parameters to local geological conditions, the filter can aggressively remove noise in smooth regions while maintaining sharp edges in complex regions, resolving the contradiction between noise reduction and edge preservation
2Object-affected harmful factors
If 3D norm zero optimization is applied to smooth geophysical data, then noise is removed and edges are preserved, but computational complexity increases
Solution Approach 1:
The patent divides the 3D geophysical data into smaller sub-volumes or blocks that can be processed independently or in parallel. This segmentation reduces the computational burden of the norm zero optimization by breaking down the large-scale complex calculation into multiple smaller, more manageable computations, thus reducing overall computational complexity while maintaining the noise removal and edge preservation benefits
Solution Approach 2:
The patent applies preliminary filtering or preprocessing steps before the main 3D norm zero optimization process. This preliminary action removes obvious noise or prepares the data in a way that reduces the computational burden of the subsequent optimization, making the overall process more efficient while still achieving the desired noise removal and edge preservation
3Measurement precision
If geophysical data is processed in full 3D volume, then comprehensive subsurface modeling is achieved, but memory requirements and processing time increase
Solution Approach 1:
The patent segments the full 3D geophysical data volume into smaller sub-volumes that can be processed independently. Each sub-volume is processed through the norm zero optimization to produce smoothed data, and these results are then assembled to form the complete subsurface model. This approach maintains comprehensive modeling accuracy while significantly reducing memory requirements and enabling parallel processing to reduce overall processing time
Data Source
AI summary
The present disclosure describes methods and systems, including computer-implemented methods, computer program products, and computer systems, for processing geophysical data. One computer-implemented method includes obtaining a set of raw geophysical data, wherein the raw geophysical data include 3-Dimensional (3D) coordinates; grouping, by a data processing apparatus, the set of the raw geophysical data into a plurality of subsets; and processing, by the data processing apparatus, each subset of the raw geophysical data using a 3D norm zero objective energy function to generate a subset of smoothed geophysical data, wherein the smoothed geophysical data is used to build a subsurface model.