Multimodal AI Rock Core Imaging for Reservoir Zone Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The manual and time-consuming process of determining oil reservoir zones from rock core images, which involves interpreting visible and ultraviolet light images and tomography, is prone to human error and lacks standardization.
Innovation Solution
A method using artificial intelligence to analyze rock core images under visible, ultraviolet, and tomography conditions, involving pre-processing, pixel and row classification, and metadata extraction to automate the classification of reservoir and non-reservoir zones.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual visual interpretation by specialist geologists is used, then interpretation accuracy can be maintained through expert judgment, but the process becomes extremely time-consuming and prone to human perception variability
Solution Approach 1:
The patent replaces the manual mechanical process of visual inspection by geologists with an automated image processing system using machine learning algorithms. The system processes visible light, ultraviolet, and tomography images through computational algorithms that classify rock zones automatically, eliminating the need for manual visual interpretation while maintaining consistency and reducing time from weeks to minutes.
Solution Approach 2:
The patent creates a digital replica of the expert geologist's interpretation process through machine learning models trained on annotated images. The system learns from examples of manually interpreted images and reproduces the classification logic algorithmically, allowing automated decision-making that mirrors expert judgment without requiring human experts for each analysis.
2Adaptability or versatility
If manual interpretation is used, then flexibility in handling complex mineralogical variations can be applied, but the process lacks standardization and is susceptible to subjective human perception
Solution Approach 1:
The patent transforms the subjective parameters of human perception into objective computational parameters. The system uses quantifiable features from multiple image types (visible, UV, tomography) and applies consistent classification thresholds and algorithms across all samples, eliminating variability in human judgment while maintaining the ability to handle complex mineralogical variations through multi-parameter analysis.
Solution Approach 2:
The patent creates a universal analysis system that handles multiple image types and rock formations through a single integrated machine learning platform. The system processes visible light, ultraviolet, and tomography images using the same algorithmic framework, providing standardized results across diverse geological conditions without requiring separate manual interpretation protocols for each case.
3Measurement precision
If multiple types of images (visible light, ultraviolet, tomography) are analyzed manually, then comprehensive characterization can be achieved, but the complexity and time required increase significantly
Solution Approach 1:
The patent merges multiple image processing streams into a unified classification system. The machine learning model simultaneously processes visible light, ultraviolet, and tomography images, integrating information from all three modalities to produce a single comprehensive classification result. This combination approach maintains high precision by utilizing complementary information from each image type while the automated system manages the complexity that would arise from analyzing multiple separate datasets.
Data Source
AI summary
The invention comprises a method for a fast, accurate and automatic determination of reservoir and non-reservoir zones of rock cores. The method involves performing pre-processing on images taken under visible light, UV light and tomography of rock cores to extract metadata therefrom and prepare them to be analyzed by a trained artificial intelligence (AI). AI performs an analysis by pixel and by row to determine reservoir and non-reservoir zones. The method further comprises filtering the AI results to reclassify relatively small regions.