Borehole Image Feature Detection Using Synthetic AI Training
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing borehole image analysis methods rely on manual dip picking, which is time-consuming and inconsistent, and current automatic methods lack consistent feature extraction and characterization, leading to inefficiencies and user bias.
Innovation Solution
A feature detection model trained with synthetic datasets is used to systematically identify and characterize geological features in borehole images, employing artificial intelligence and iterative updates based on loss functions to enhance prediction confidence.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual dip picking is used for feature identification, then measurement precision can be maintained through expert judgment, but productivity is significantly reduced due to time-consuming analysis
Solution Approach 1:
The patent replaces manual mechanical analysis (expert dip picking) with an automated computer-based feature detection model that uses machine learning to identify geological features in borehole images, thereby maintaining measurement precision while dramatically improving productivity
Solution Approach 2:
The patent creates synthetic borehole image datasets that replicate real geological features to train the feature detection model, allowing the system to learn from numerous examples without requiring manual analysis of each actual image, thus resolving the contradiction between accuracy and speed
2Productivity
If automatic feature detection methods are implemented, then productivity is improved through automated analysis, but measurement precision deteriorates due to inconsistent feature extraction
Solution Approach 1:
The patent performs preliminary training of the feature detection model using extensive synthetic datasets before actual feature detection, preparing the system in advance to consistently identify features across different borehole images, thereby ensuring both productivity and measurement precision
Solution Approach 2:
The patent implements iterative training where the model receives feedback from loss function calculations comparing predicted features with ground truth labels, continuously improving feature extraction consistency while maintaining automated high-speed analysis
3Adaptability or versatility
If synthetic datasets are used for model training, then adaptability is improved through comprehensive feature coverage, but device complexity increases due to data generation and model training requirements
Solution Approach 1:
The patent varies parameters in synthetic borehole image generation (such as feature orientations, depths, and geological conditions) to create diverse training examples, improving model adaptability while using computational algorithms rather than physical equipment to avoid increasing device complexity
Data Source
AI summary
A system is provided that includes a processing circuitry and a memory, accessible by the processing circuitry, the memory storing instructions that, when executed by the processing circuitry cause the processing circuitry to perform operations. The operations may include generating synthetic dataset representative of a plurality of synthetic borehole images and training a feature detection based on the synthetic dataset, wherein the feature detection model predicts one or more features associated with a plurality of borehole images at one or more depths, one or more parameters associated with the one or more features, or both. The operations may also include generating a prediction dataset comprising a predicted detection vector and a predicted parameter matrix based on the feature detection model.