Microstructure Image Analysis Using Ideal Convergence Distance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for evaluating dispersibility in microstructure images are limited by restrictions on compartment division, failing to account for positional relations and aspect ratios, making precise quantification of dispersibility and heterogeneity difficult.
Innovation Solution
An image analysis device and method that calculates a heterogeneity index based on ideal convergence distance and connection distance correlation information, allowing for precise evaluation of object arrangement bias in microstructure images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If compartmentalization method is used to evaluate dispersibility, then dispersibility can be quantified, but positional relations of objects are not expressed and analysis is restricted by image size and aspect ratio
Solution Approach 1:
The patent segments the image analysis into multiple scales by dividing the image into compartments of different sizes (2^k × 2^k pixels). This allows the method to analyze dispersibility at various resolution levels, making it adaptable to images of any size while maintaining the ability to express positional relations through hierarchical compartment division.
Solution Approach 2:
The patent introduces a new dimension of analysis by considering multiple scale levels (k = 0, 1, 2, ...) simultaneously. Instead of analyzing only at a single fixed scale, the method evaluates dispersibility across a spectrum of compartment sizes, adding the scale dimension to the traditional binary partitioning approach and enabling flexible adaptation to different image dimensions.
2Ease of operation
If compartment size is made constant for simplicity, then calculation is easier, but positional relation information is lost
Solution Approach 1:
The patent makes the compartment size dynamic by allowing it to vary across different scales (2^k × 2^k pixels for k = 0, 1, 2, ...). This dynamic adjustment of compartment size enables the method to capture positional relations at multiple resolutions while maintaining computational tractability through systematic scaling rather than fixed rigid partitions.
Solution Approach 2:
The patent performs preliminary division of the image into hierarchical compartments before conducting the dispersibility analysis. By pre-establishing the multi-scale compartment structure, the method prepares the image data in advance to preserve positional relation information across different scales, avoiding information loss during the subsequent analysis steps.
3Stability of the object's composition
If division number is restricted to 2^n × 2^n for constant compartment size, then compartment size remains constant, but analysis cannot be performed on images with certain pixel sizes and aspect ratios
Solution Approach 1:
The patent changes the parameter of compartment size from fixed to variable by introducing a scale parameter k that can take multiple values (0, 1, 2, ...). This allows the compartment size to be adjusted to 2^k × 2^k pixels, enabling the method to handle images of various dimensions and aspect ratios while maintaining systematic compartment division through power-of-2 scaling.
Data Source
AI summary
An image analysis device analyzes an arrangement of objects in a microstructure image of a composition and includes an ideal convergence distance calculation unit configured to calculate an ideal convergence distance indicating a distance between the objects in an ideal state of the arrangement of the objects. A convergence calculation processing unit is configured to generate, for each of a plurality of connection distances, a plurality of remaining groups in each of which the objects are connectable to each other at the connection distance, and connection distance correlation information is generated indicating a correlation between a group number of the generated remaining groups and the connection distance. A heterogeneity index is calculated indicating a bias in the arrangement of the objects based on the ideal convergence distance and the connection distance correlation information; and display data is output and displayed according to the heterogeneity index.