Rock Core Image Segmentation for Accurate Facies Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for geological rock core analysis are manual and prone to human error and subjective interpretation, lacking efficiency and accuracy in identifying geological features and predicting missing sections.
Innovation Solution
A method utilizing a U-Net segmentation model for image processing, including supervised mask generation, image tiling, model training, and output processing to generate a stacked color image with synthetic rock sections and facies classifications, enhancing the analysis of geological rock cores.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual methods are used for geological rock core analysis, then human interpretation and identification can be performed, but the process is prone to human error and subjective interpretation, lacking efficiency and accuracy
Solution Approach 1:
The patent replaces manual mechanical analysis methods with an automated image processing system using machine learning models. The U-Net segmentation model automatically processes rock core images to identify geological features, eliminating human subjectivity and error while maintaining high accuracy through algorithmic consistency and reproducibility
Solution Approach 2:
The system enables self-service analysis where the machine learning model autonomously processes images without continuous human intervention. The automated pipeline includes independent steps for image preprocessing, segmentation, feature extraction, and classification, allowing the system to perform complete geological analysis independently while improving both accuracy and efficiency
2Productivity
If manual analysis methods are used, then flexibility in interpretation is maintained, but the process lacks efficiency and is time-consuming
Solution Approach 1:
The system performs preliminary actions by pre-processing images and pre-training machine learning models on extensive datasets before actual analysis. The U-Net model is pre-trained to recognize geological patterns, enabling rapid processing of new images without requiring time-consuming manual interpretation for each sample
Solution Approach 2:
Manual analysis processes are replaced with automated computational systems that process images through standardized algorithms. The machine learning pipeline automatically executes segmentation, feature detection, and classification tasks that would otherwise require significant manual time investment, dramatically improving efficiency while reducing analysis time
3Measurement precision
If comprehensive image processing is performed on entire rock core images, then complete analysis is achieved, but computational complexity and processing time increase
Solution Approach 1:
The patent applies segmentation by dividing the rock core images into smaller manageable segments or patches for processing. The U-Net model processes images in segmented regions, identifying geological features in each segment independently, then combines results to achieve complete coverage. This reduces computational complexity while maintaining comprehensive analysis through systematic division of the processing task
Solution Approach 2:
The system employs partial action by initially focusing on processing and identifying key geological features in representative samples or specific regions of interest. The machine learning model is trained on partial datasets and then applied to broader image sets, achieving comprehensive analysis through iterative processing that balances completeness with computational feasibility
4Measurement precision
If automated machine learning methods are used, then accuracy and efficiency are improved, but the system complexity increases
Solution Approach 1:
The patent implements a universal machine learning framework where the U-Net segmentation model serves multiple functions: image preprocessing, feature segmentation, classification, and prediction of geological properties. This multi-functional approach consolidates what would otherwise require multiple separate systems into a single integrated platform, improving accuracy while managing complexity through functional consolidation
Solution Approach 2:
The system introduces intermediary components such as data preprocessing layers, feature extraction modules, and post-processing algorithms that mediate between raw images and final interpretations. These intermediary elements simplify the core machine learning task while maintaining high accuracy, effectively managing system complexity through modular architecture that separates concerns across multiple processing stages
Data Source
AI summary
A method includes: providing color images for a geological rock core sample, creating initial masks for a subset of the color images, dividing each color image into sets of image tiles, splitting the sets of image tiles into a training set and a validation set, augmenting the training and validation sets, including orienting each image tile in a same direction by sample depth, training a model with the augmented training validation sets, generating image masks from the trained model, corresponding to the color images, combining the sets of image tiles to regenerate each of the color images from their image tile sets, applying the generated image masks to the color images to generate greyscale masked images, stacking the greyscale masked images by sample depth, and applying colors to the stacked greyscale masked images to generate and display a stacked color image representing the entire sample length.