SEM Mask Inspection Image Stitching via Overlap Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current mask inspection apparatuses using scanning electron microscopes face challenges in generating high-definition images with a wide field of view due to limited field of view and significant aberrations, requiring extensive correction efforts and time-consuming image patching processes.
Innovation Solution
A mask inspection apparatus and method that divides the observed area into overlapping sections, allowing for accurate alignment and combination of adjacent images to form a complete SEM image with a wide field of view and high definition, using irradiation, detection, image processing, and control means to correct coordinate data and combine images based on overlap areas.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If a wide field of view is used to capture the entire photomask pattern, then the field of view increases, but image definition deteriorates due to aberrations such as astigmatism, field curvature and distortion
Solution Approach 1:
The patent divides the wide field of view into multiple smaller sub-fields, each captured with high definition by the SEM. Multiple SEM images are acquired for different sub-fields and then synthesized into a single wide-field high-definition image, resolving the contradiction between wide field of view and image definition.
2Area of stationary object
If multiple divided SEM images are synthesized to achieve wide field of view, then the field of view increases, but the complexity of aberration correction increases significantly
Solution Approach 1:
By segmenting the wide field into multiple sub-fields that can be captured individually with high definition, the patent avoids the need for complex real-time aberration correction across the entire wide field. Each sub-field is corrected independently or captured with minimal aberration.
Solution Approach 2:
The patent creates a synthesized copy of the wide-field image by combining multiple high-definition sub-field images. This synthetic approach allows wide field of view without requiring the physical optical system to maintain high definition across the entire wide field simultaneously.
3Area of stationary object
If multiple divided SEM images are synthesized manually to achieve wide field of view, then the field of view increases, but the time required for image processing increases significantly
Solution Approach 1:
The patent implements automated image synthesis where the system itself performs the alignment and combination of multiple divided SEM images using coordinate data and overlap area analysis. This self-service automation eliminates manual processing time while achieving wide field of view synthesis.
Solution Approach 2:
The patent performs preliminary acquisition of coordinate data for each divided image during the imaging process itself. This preliminary preparation enables rapid automated synthesis later, reducing the overall time loss by having alignment information ready before the synthesis operation begins.
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
Enables the rapid creation of high-definition SEM images with a wide field of view, reducing aberration correction loads and improving inspection accuracy by automating the image combination process.
Implementation Method 1
a sample is irradiated with incident electrons while the surface of the sample in an electron beam scanning region is scanned by the incident electrons, and secondary electrons emitted from the sample are acquired through a scintillator
Data Source
AI summary
A mask inspection system includes irradiation means for irradiating a sample with an electron beam, electron detection means for detecting a quantity of electrons generated from the sample, image processing means, storage means, and control means for determining divided areas in such a way that divided images adjacent to each other overlap with each other, and acquiring the divided images of the respective divided areas. The control means extracts two divided images adjacent to each other in a predetermined sequence, then detects an image of a same pattern formation area included in an overlap area, and determines the detected image to be a combination reference image. The control means then combines the two divided images adjacent to each other on the basis of the combination reference image to thereby form an entire SEM image of the observed area.