Multi-Beam SEM Noise Mitigation via Image Comparison
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Multi-beam scanning electron microscopy systems face challenges in noise mitigation, as existing techniques are ineffective in separating noise from image information, leading to difficulties in identifying and removing noise components, especially in semiconductor device inspection where noise sources like vibrational, acoustic, and electrical noise are prevalent.
Innovation Solution
A multi-beam scanning electron microscopy system equipped with a controller that processes multiple images to identify and remove common noise components, using image comparison algorithms to correct for positional and intensity noise, and optionally rescan images to address focus errors, thereby forming aggregated images with reduced noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multi-beam SEM systems are used to reduce inspection time, then productivity is improved, but noise mitigation becomes ineffective
Solution Approach 1:
The patent segments the image processing task by dividing it into multiple passes: first identifying noise components in individual images, then systematically removing them across the multi-beam image set. This segmentation allows noise mitigation to be effectively integrated into the high-speed multi-beam inspection workflow
Solution Approach 2:
The system implements feedback by using identified noise components from initial image analysis to guide subsequent noise removal operations. The noise characteristics detected in early processing stages are fed back into the image correction algorithm, enabling continuous refinement of noise mitigation while maintaining inspection speed
2Measurement precision
If existing noise mitigation techniques are applied to multi-beam SEM images, then image quality improvement is attempted, but the techniques prove ineffective
Solution Approach 1:
The patent employs dynamic adaptation by adjusting noise mitigation parameters based on the specific characteristics of each multi-beam image set. The system dynamically identifies noise components and adjusts removal strategies according to the actual noise patterns present, rather than applying static conventional filters
Solution Approach 2:
The system changes processing parameters adaptively by modifying noise identification and removal parameters based on the specific multi-beam image characteristics. This includes adjusting sensitivity thresholds and processing intensity based on the detected noise levels and patterns in each inspection scenario
3Measurement precision
If conventional image processing is used in single beam SEM, then noise separation is attempted, but insufficient feature information prevents effective noise extraction
Solution Approach 1:
The patent merges information from multiple multi-beam images to compensate for insufficient feature information in individual images. By combining data across the beam array, the system accumulates sufficient feature information to enable effective noise extraction while preserving legitimate image features
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system effectively reduces common mode noise in multi-beam SEM systems, improving image quality by distinguishing noise from actual features and reducing false defect identification, enhancing inspection efficiency in semiconductor device fabrication.
Implementation Method 1
a multi-beam electron beam source configured to generate a plurality of electron beams
Implementation Method 2
a detector assembly configured to detect a plurality of electron signal beams emanating from the surface of the sample to form a plurality of images
Data Source
AI summary
A scanning electron microscopy system is disclosed. The system includes a multi-beam scanning electron microscopy (SEM) sub-system. The SEM sub-system includes a multi-beam electron beam source configured to generate a plurality of electron beams, a sample stage configured to secure a sample, an electron-optical assembly, and a detector assembly configured to detect a plurality of electron signal beams emanating from the surface of the sample to form a plurality of images, each image associated with an electron beam of the plurality of electron beams. The system includes a controller configured to receive the images from the detector assembly, compare two or more of the images to identify common noise components present in the two or more images, and remove the identified common noise components from one or more images of the plurality of images.