Automated Precipitate Detection in Alloy Electron Microscopy Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting precipitates in alloys using transmission electron microscopy images are manual, time-consuming, and prone to human error due to inhomogeneous backgrounds, making automated image processing challenging, especially in high-magnification or dark-field images.
Innovation Solution
A method utilizing local image processing techniques, including the search for light peaks, application of approximation functions, and determination of potential event lengths to identify precipitates, which are then filtered and classified based on predetermined criteria, enabling automated detection despite image inhomogeneities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual image analysis is used to detect precipitates, then detection reliability is improved, but productivity deteriorates
Solution Approach 1:
The system performs self-service by automatically detecting precipitates through image processing algorithms. The computer-implemented method autonomously identifies potential events, determines their dimensions, and classifies them as precipitates without requiring manual intervention, thereby maintaining detection reliability while significantly improving productivity
Solution Approach 2:
The manual mechanical process of visual inspection and manual measurement is replaced by an automated image processing system. The system uses digital image analysis, approximation functions, and automated dimension determination to substitute the manual mechanical approach, achieving both high reliability and improved productivity
2Productivity
If automated image processing is applied, then productivity is improved, but measurement precision deteriorates due to inhomogeneous backgrounds
Solution Approach 1:
The system applies local quality by determining the dimensions of potential events based on local image characteristics around each event. The approximation function is applied locally to each potential event to determine its dimensions, adapting to local variations in background intensity and texture, thereby maintaining measurement precision while enabling automated processing
Solution Approach 2:
The system changes parameters by using approximation functions to model the intensity profile around potential events. By fitting mathematical models to the local image data and extracting dimensional parameters from these models, the system achieves precise measurements despite variations in background conditions, enabling reliable automated detection
3Ease of operation
If traditional thresholding is used for automated detection, then ease of operation is improved, but reliability deteriorates due to background inhomogeneity
Solution Approach 1:
The simple but unreliable traditional thresholding method is replaced by a more sophisticated automated image processing system. The system uses approximation functions and local dimension determination to substitute the simple thresholding approach, maintaining ease of automated operation while significantly improving reliability by accurately distinguishing precipitates from background variations
Data Source
Figure 1
Figure 2~3
AI summary
The invention relates to the field of material characterization by image analysis. It concerns a method and a device for detecting precipitates in an alloy by analyzing an image obtained by electron microscopy, and a method for determining a mechanical property of an alloy.According to the invention, the method (13) comprises: ▪ for each row of pixels of the image, a search (132) for light peaks, each peak being associated with a potential event, ▪ for each light peak: ∘ an application (133) of an approximation function, the approximation function approximating the gray levels of pixels of the line considered around said light peak, ∘ a determination (134) of a potential event length (Lep), and ∘ a comparison (135) of the potential event length (Lep) to a predetermined minimum length (Lmin), the potential event being considered as a precipitate of the alloy if its length is greater than or equal to the predetermined minimum length, and being ignored otherwise.