Wafer Noise Defect Extraction Using Histogram Similarity Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting and removing noise defects on semiconductor wafers require precise parameter settings, which are difficult to determine due to the complex and irregular patterns of wafer defects, making it challenging to accurately classify defects.
Innovation Solution
A method that extracts noise defects by randomly selecting abnormal value calculators, calculating abnormal values, generating frequency histograms, and clustering defects using spectral clustering without setting specific parameters, utilizing various abnormal value calculators such as AggkNN, CCPOD, and others.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If distance-based or density-based parameters are precisely set to increase detection accuracy, then noise defect detection accuracy is improved, but the complexity of parameter setting increases significantly
Solution Approach 1:
The system performs self-service by automatically determining appropriate distance-based and density-based parameters through iterative calculation and evaluation, eliminating the need for manual parameter setting by experts. The defect management apparatus autonomously adjusts parameters based on defect pattern analysis, making the complex parameter setting process transparent to users.
Solution Approach 2:
The system dynamically changes parameters by iteratively adjusting distance-based parameters (e.g., defect spacing thresholds) and density-based parameters (e.g., defects per unit area) to optimize noise defect detection. Multiple parameter sets are evaluated to find the most appropriate configuration for different defect patterns.
2Reliability
If multiple abnormal value calculators are used to improve noise defect extraction accuracy, then detection reliability is improved, but computational complexity increases
Solution Approach 1:
The system merges multiple abnormal value calculators (e.g., distance-based calculators, density-based calculators, pattern recognition calculators) into a unified defect management apparatus. These calculators work together synergistically, with each contributing different aspects of defect analysis, to improve overall noise defect extraction reliability.
Solution Approach 2:
The defect management apparatus is segmented into multiple specialized abnormal value calculators, each responsible for specific calculation tasks (e.g., spatial distance calculation, density calculation, pattern matching). This segmentation allows parallel processing and reduces the computational burden on any single calculator while maintaining high accuracy.
Data Source
AI summary
There is provided a method of extracting a noise defect from defect data. The method comprises obtaining the defect data including data on a plurality of defects, determining one of a plurality of abnormal value calculators for calculating abnormal values using each of a variety of predetermined calculation methods and calculating an abnormal value for each of the plurality of defects using one of the plurality of abnormal value calculator, generating a frequency histogram for each of the plurality of defects using the abnormal value, calculating similarity between the plurality of defects using the frequency histogram for each of the plurality of defects, and extracting the noise defect from the plurality of defects based on the similarity