3D Rock Image Segmentation for Consistent Reservoir Characterization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The process of segmenting rock fabrics for reservoir characterization is time-intensive, memory-intensive, and prone to inconsistent results due to manual methods, making it challenging to maintain consistency and efficiency in rock image analysis.
Innovation Solution
A machine learning-based workflow for three-dimensional (3D) segmentation of rock images, involving resizing, partitioning into cubes, extracting orthogonal planes, and clustering features to enhance accuracy and consistency, while reducing computational time and memory usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual methods are used for segmenting rock fabrics, then flexibility and adaptability are maintained, but the process becomes time-intensive and memory-intensive
Solution Approach 1:
The patent replaces manual mechanical segmentation methods with an automated machine learning-based workflow. The system uses trained models to automatically segment rock fabrics in 3D images, eliminating the need for manual intervention while significantly reducing processing time and memory requirements.
Solution Approach 2:
The patent changes the processing approach by working with resized images at reduced resolution during segmentation, then mapping results back to original dimensions. This parameter change in image resolution dramatically reduces memory usage and computational time while maintaining segmentation accuracy through the mapping process.
2Measurement precision
If manual segmentation methods are used, then detailed analysis is possible, but consistency and efficiency deteriorate
Solution Approach 1:
The patent replaces manual segmentation with automated machine learning models that provide consistent, repeatable results. The trained models apply the same segmentation criteria uniformly across all images, eliminating variability inherent in manual methods while maintaining high accuracy through sophisticated feature extraction and classification algorithms.
3Manufacturing precision
If full-resolution 3D images are processed directly, then segmentation detail is maximized, but memory usage and computational time increase significantly
Solution Approach 1:
The patent temporarily changes the resolution parameter by processing images at a reduced size during the segmentation phase. This dramatically reduces memory requirements and computational load. The system then maps the segmentation results back to the original full-resolution images, preserving detailed fabric characterization without requiring high memory usage during processing.
Solution Approach 2:
The patent divides the 3D image processing into distinct segments: resizing the image, performing segmentation on the resized image, extracting features, and finally mapping results back to original dimensions. This segmented approach allows efficient processing at reduced resolution while maintaining the ability to analyze detailed fabric characteristics in the original image space.
Data Source
AI summary
Systems and methods are provided for determining fabrics of a geological sample using three-dimensional segmentation. An example method can include receiving three-dimensional (3D) image of a geological sample, adjusting an initial size of the 3D image of the geological sample, and partitioning the resized 3D image of the geological sample into cubes. The example method can include, for each cube, generating orthogonal planes based on a center of mass of each cube and extracting, for the orthogonal planes associated with each cube, one or more features to represent texture of the geological sample. The example method can further include grouping the cubes into one or more clusters based on the one or more features and constructing a volume of the resized 3D image of the geological sample based on the one or more clusters for a texture analysis of the geological sample.