SEM Image Alignment Using GAN Conversion for Error Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing SEM equipment inspection methods struggle with accurately comparing and aligning heterogeneous images, leading to measurement errors and time-consuming manual corrections.
Innovation Solution
A method involving the acquisition of SEM and design images, pre-processing, training with a Generative Adversarial Network (GAN) algorithm to generate a conversion model, converting SEM images into design-like images, and extracting alignment coordinates to determine and correct measurement errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If existing inspection methods are used to compare and align SEM images with design images, then the process can be automated to some extent, but the alignment accuracy deteriorates when images are heterogeneous, leading to inspection failures
Solution Approach 1:
The patent introduces a conversion model as an intermediary that transforms SEM images into design image format. This mediator enables accurate comparison between heterogeneous images by converting them to a common representation, resolving the alignment accuracy issue while maintaining automation
Solution Approach 2:
The patent creates a converted copy of the SEM image that mimics the design image format. By generating this copied version through the conversion model, the system can accurately compare and align images without manual intervention, solving both automation and precision requirements
2Measurement precision
If manual comparison and alignment are performed to achieve accurate results, then measurement precision improves, but the process becomes time-consuming and expensive
Solution Approach 1:
The system performs self-service by automatically generating the conversion model and using it to align images without human intervention. The automated conversion and alignment process achieves manual-level accuracy while eliminating time loss and cost associated with manual operations
Solution Approach 2:
The patent changes the parameter representation of SEM images by converting them to design image format through the conversion model. This parameter transformation enables automated processing to achieve accuracy previously only attainable through manual methods, reducing time and cost
3Device complexity
If traditional image comparison methods are used, then the process remains simple, but the ability to handle heterogeneous images deteriorates, causing inspection failures
Solution Approach 1:
The conversion model provides universality by transforming SEM images into a format compatible with design images. This multi-functional approach allows the system to handle heterogeneous images effectively while maintaining a relatively simple comparison process, improving adaptability without significantly increasing complexity
Data Source
AI summary
There are provided a method of accurately detecting a measurement error of SEM equipment by comparing and aligning a design image with an SEM image, and a method of accurately aligning SEM equipment based on a detected measurement error. The method of detecting a measurement error of SEM equipment includes acquiring SEM images of a measurement target, performing pre-processing on the SEM images and design images corresponding thereto, selecting training SEM images from among the SEM images, performing training by using the training SEM images and training design images and generating a conversion model between the SEM images and the design images, converting the SEM images into conversion design images by using the conversion model, extracting an alignment coordinate value by comparing and aligning the conversion design images with the design images, and determining a measurement error of the SEM equipment based on the alignment coordinate value.