SEM Pattern Defect Classification Using Merged Color Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for analyzing pattern defects in semiconductor manufacturing require manual die-to-die SEM analysis, which is time-consuming, and machine learning approaches need prior learning of various defect shapes, making them inefficient.
Innovation Solution
A method involving alignment and merging of SEM images with different colors to locate and classify defects by color difference, followed by comparing with a pattern layout to determine defect type.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual die-to-die SEM analysis is used, then measurement precision is maintained, but productivity is reduced and loss of time increases
Solution Approach 1:
The patent creates a merged SEM image that combines multiple individual SEM images into a single composite representation. This copied/merged image serves as a comprehensive reference that enables automated defect detection and classification, replacing manual analysis while maintaining precision through the integrated visual information from multiple sources
Solution Approach 2:
The merged SEM image acts as an intermediary between multiple individual SEM images and the defect classification system. By combining multiple images into one standardized format with consistent coloring and alignment, it enables automated algorithms to efficiently process and classify defects without requiring manual intervention in each individual image
2Productivity
If machine learning method is used, then productivity is improved, but device complexity increases due to requirement of learning various defect shapes in advance
Solution Approach 1:
The patent performs preliminary actions by merging and aligning multiple SEM images into a single standardized merged image before defect classification. This preprocessing step creates a unified reference format that simplifies subsequent automated analysis, eliminating the need for complex machine learning models to learn various defect shapes from scratch
Solution Approach 2:
The merged SEM image serves as a universal reference that can be used for classifying multiple types of defects across different patterns. By creating a single standardized image that integrates information from multiple sources, the system achieves multi-functionality in defect detection without requiring separate learning models for each defect type
3Measurement precision
If multiple SEM images are merged with different colors, then measurement precision is improved for defect location, but device complexity increases
Solution Approach 1:
The patent segments the defect analysis process into distinct steps: merging multiple SEM images with different color representations, aligning them to a reference pattern, and then classifying defects based on color differences. This segmentation of the image processing workflow enables precise defect location through color comparison while managing complexity through systematic step-by-step processing
Data Source
AI summary
In a method of classifying defects of a pattern, SEM images having different colors on the same pattern are aligned and merged. The defects of the pattern are located by color difference in the merged SEM image. The defects of the pattern are firstly located at the location of the defects. A design in which a layout of the pattern is drawn, and the merged SEM image are aligned. The defects of the pattern are secondly classified by comparing the layout of the pattern drawn in the design and the merged SEM image. A type of the defects of the pattern is determined by combination of the first and second classifications of the defects of the pattern.