Inversion Image Boundary Detection for Subsurface Feature Steering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing drilling systems face challenges in accurately determining the location, shape, orientation, and boundaries of subsurface features due to difficulties in interpreting measurement data.
Innovation Solution
A boundary detection system utilizing a subsurface boundary machine learning model processes individual pixels of inversion images to generate boundary masks, enabling efficient and accurate identification of subsurface features.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional measurement data interpretation methods are used to determine subsurface features, then comprehensive analysis can be performed, but accuracy in determining location, shape, orientation and boundaries is insufficient
Solution Approach 1:
The patent replaces traditional mechanical interpretation methods with a machine learning model that processes inversion images to identify subsurface feature boundaries. The system uses trained neural networks to automatically detect boundaries, replacing manual or conventional algorithmic analysis with intelligent automated recognition, thereby improving measurement precision without proportionally increasing system complexity
Solution Approach 2:
The patent introduces inversion images as an intermediary representation between raw measurement data and boundary identification. The machine learning model operates on these processed images rather than raw data, creating an intermediate layer that simplifies the detection task and improves boundary identification accuracy while managing computational complexity
2Measurement precision
If detailed analysis of subsurface features is performed, then better understanding of underground structures is achieved, but time and computational resources increase
Solution Approach 1:
The patent performs preliminary processing by generating inversion images from raw measurement data before applying the machine learning model. This pre-processing step organizes the data into a format optimized for rapid boundary detection, reducing the computational burden during actual analysis and decreasing processing latency while maintaining high identification accuracy
Solution Approach 2:
The patent segments the analysis process into distinct stages: data acquisition, inversion image generation, and boundary detection using machine learning. This segmentation allows each stage to be optimized independently, with the machine learning model focusing specifically on boundary identification tasks, thereby reducing overall processing time while maintaining precision
Data Source
AI summary
A method of identifying subterranean features includes receiving an inversion image indicating a portion of a subsurface feature. Boundary information is determined for the inversion images using a subsurface boundary machine learning model that is generated to process individual pixels of input inversion images through a decision-based architecture to identify boundaries of subsurface features. Based on the boundary information, a boundary mask is generated for the inversion image. The method further includes providing the boundary mask for adjusting one or more downhole parameters based on the boundary mask.