Learning-Type Classifying Apparatus Region Integration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing learning-type classifying apparatuses face challenges in accurately classifying defective regions in images and creating effective teacher data, particularly in integrating regions and correcting classification errors, which affects the efficiency and accuracy of defect classification.
Innovation Solution
A learning-type classifying apparatus comprising a region extracting unit, characteristic value calculating unit, classifying unit, region integrating unit, display unit, input unit, and teacher data creating unit, which allows for the extraction, classification, integration, and correction of regions, as well as the creation of teacher data, with the ability to recluster and reintegrate regions based on user input and specified characteristic axes, and judges the necessity of additional teacher data based on classification correctness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If regions are integrated based on classification results, then the number of regions to be processed is reduced, but classification accuracy may deteriorate due to erroneous integration of regions belonging to different classes
Solution Approach 1:
The system displays integrated regions to the user and receives feedback on whether the integration is correct. Based on this feedback, the system learns from errors and improves future integration decisions, resolving the contradiction between automated processing efficiency and classification accuracy.
Solution Approach 2:
The system automatically creates teacher data from user corrections and uses it to improve its own classification and integration performance without requiring manual reprogramming, enabling the system to self-correct integration errors while maintaining high processing efficiency.
2Manufacturing precision
If manual correction of classification errors is performed for each region, then classification accuracy is improved, but the time and labor required increases significantly
Solution Approach 1:
The system merges multiple individual region corrections into a single integrated region correction. When a user corrects an integrated region, the correction is automatically applied to all constituent regions, reducing correction time while maintaining accuracy.
Solution Approach 2:
The system creates teacher data by copying the corrected classification results from integrated regions to individual regions. This automatic copying eliminates the need for manual correction of each region while ensuring consistent accuracy across all regions.
3Manufacturing precision
If the number of teacher data is increased to improve classification correctness, then classification accuracy is improved, but the complexity of data management and processing increases
Solution Approach 1:
The system performs preliminary integration of regions and creates integrated teacher data before detailed classification. This preliminary organization reduces the total number of individual teacher data entries needed while maintaining or improving classification correctness.
Solution Approach 2:
The system uses integrated regions as multi-functional units that serve both as classification targets and as sources for generating teacher data for multiple individual regions. This universal approach reduces data management complexity while improving classification accuracy.
Data Source
AI summary
A learning-type classifying apparatus comprises defective region extracting unit for extracting defective regions of classification targets from an image in which the plurality of regions of the classification targets are present, characteristic value calculating unit for calculating characteristic values for the extracted regions of the classification targets, classifying unit for classifying the extracted regions of the classification targets into predetermined classes on the basis of the calculated characteristic values, region integrating unit for integrating the regions which belong to the same class as a result of the classification, display unit for displaying images of the integrated regions and the classification results, input unit for correcting errors in the classification results, and teacher data creating unit for creating teacher data for each of the regions so that the classification results of the integrated regions are reflected in each region included in the integrated regions.