SEM Image Alignment Auto Tuning for Faster Accurate Inspection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current manual methods for aligning SEM images with design layout data are time-consuming and tedious, often requiring repetitive trial-and-error processes, especially for challenging alignment cases.
Innovation Solution
A result-oriented auto parameter tuning method is introduced, which involves acquiring an inspection image and a reference image, determining a target alignment, evaluating alignment parameter combinations, selecting the optimal combination, and applying it to the reference image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual trial-and-error methods are used for alignment parameter tuning, then alignment accuracy can be achieved, but the time required for alignment becomes excessively long
Solution Approach 1:
The system performs self-alignment by automatically evaluating multiple alignment parameter combinations and selecting the optimal one based on image correlation metrics, eliminating the need for manual trial-and-error adjustment while achieving accurate alignment between inspection images and reference images
Solution Approach 2:
The system implements a feedback mechanism where alignment parameter combinations are evaluated based on their ability to align inspection image patterns with reference image patterns, and the evaluation results are used to select and apply the optimal alignment parameters, creating a closed-loop automatic tuning process
2Productivity
If automatic parameter tuning is implemented, then alignment time is significantly reduced, but the complexity of the alignment system increases
Solution Approach 1:
The system creates a digital model of the alignment evaluation process by generating synthetic reference images with known alignment parameters and comparing them against actual inspection images, allowing automatic determination of optimal alignment parameters through pattern recognition and correlation analysis without requiring complex physical adjustment mechanisms
Data Source
AI summary
An improved method and system for image alignment of an inspection image are disclosed. An improved method comprises acquiring an inspection image, acquiring a reference image corresponding to the inspection image, acquiring a target alignment between the inspection image and the reference image based on characteristics of the inspection image and the reference image, estimating an alignment parameter based on the target alignment, and applying the alignment parameter to a subsequent inspection image.