Borehole Image Blending via Supervised Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The manual application of image blending techniques for electromagnetic imager tools is time-consuming and prone to inconsistencies, requiring numerous iterations by experienced operators to adjust various parameters, leading to inconsistent image quality across different wellbores and zones.
Innovation Solution
A supervised machine learning technique is employed to train a blending parameter machine learning model, which identifies optimal blending parameters for image blending, reducing the need for manual adjustments and improving consistency by applying these parameters to measurements made across multiple frequencies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If manual image blending technique is applied by operator, then image quality can be adjusted, but processing time increases and consistency deteriorates
Solution Approach 1:
The system performs self-service by automatically selecting optimal images and blending parameters without requiring manual operator intervention. The image blending technique is controlled autonomously by the system, which evaluates multiple images and determines the best combination to achieve consistent quality across different wellbores and zones.
Solution Approach 2:
The system changes blending parameters automatically based on the specific characteristics of each wellbore and zone. By dynamically adjusting parameters such as blending weights and selection criteria, the system maintains consistent image quality without requiring manual reconfiguration for each case, thereby reducing processing time while preserving quality consistency.
2Manufacturing precision
If multiple iterations are performed to adjust parameters, then image quality improves, but processing time and complexity increase
Solution Approach 1:
The system performs preliminary action by pre-evaluating multiple images and pre-determining the optimal blending parameters before final image generation. This upfront analysis allows the system to achieve accurate blending results in a single pass, eliminating the need for multiple iterative adjustments and thereby improving processing efficiency.
Solution Approach 2:
The system implements feedback mechanisms to evaluate the quality of blended images and automatically adjust parameters based on this evaluation. This closed-loop approach ensures high blending parameter accuracy while minimizing the number of iterations required, as the system learns from each evaluation and converges to the optimal solution more efficiently.
3Measurement precision
If experienced operator manually controls variables, then image blending accuracy improves, but ease of operation deteriorates and time consumption increases
Solution Approach 1:
The system replaces the mechanical system of manual operator control with an automated computational system. The image blending technique is controlled through algorithmic processes that automatically select images and adjust parameters, eliminating the need for manual intervention while maintaining or improving blending accuracy through systematic evaluation of multiple parameters.
Solution Approach 2:
The system achieves universality by designing a multi-functional automated system that can handle various wellbores and zones with a single unified approach. This automated system performs multiple functions including image selection, parameter optimization, and quality assessment, thereby maintaining high blending accuracy across diverse applications without requiring specialized manual control for each case.
Data Source
AI summary
Aspects of the subject technology relate to systems, methods, and computer-readable media for controlling borehole imaging blending through machine learning. A blending parameter machine learning model can be trained through a supervised machine learning technique with a dataset of known input and known output associated with an electromagnetic imager tool. The blending parameter machine learning model is associated with an image blending technique for blending images generated through the electromagnetic imager tool at different frequencies. One or more blending parameters for the image blending technique can be identified by applying the blending parameter machine learning model to measurements of the electromagnetic imager tool operating to log a wellbore across a plurality of frequencies. One or more blended images can be generated by applying the image blending technique according to the one or more blending parameters to a plurality of images of the measurements made by the electromagnetic imager tool.