Semiconductor Mask Defect Detection Using Multi-Palette Color Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional defect inspection systems face challenges in accurately detecting and visually identifying tiny defects on semiconductor masks due to the small size of LSI circuit patterns, as they require precise definition of pattern edges and differentiation between defects and pattern edges, which is difficult with conventional single-color palette representations.
Innovation Solution
The use of multiple color palettes, with different palettes for reference, optical, and comparison images, and applying gamma correction to enhance visibility, allows for two-dimensional gradation value representation, making it easier to identify and predict defect types by displaying differences in gradation values between reference and optical images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single color palette with 256 gradation values is used to display comparison images, then the display system is simple, but tiny defects become difficult to visually identify and differentiate from pattern edges
Solution Approach 1:
The single color palette is segmented into multiple color palettes, each optimized for specific display purposes. The patent divides the display system into at least two color palettes: one for displaying the optical image and another for displaying the comparison image, allowing each palette to be optimized for its specific function and improving defect visibility without overwhelming complexity
Solution Approach 2:
The patent transitions from a one-dimensional gradation value representation (0-255) to a two-dimensional gradation value representation by using multiple color palettes. This dimensional change allows the system to preserve more information about the relationship between optical image values and comparison image values, enabling better defect identification while managing complexity through structured organization
2Measurement precision
If the number of color palettes is increased to improve defect visibility, then defect detection accuracy improves, but system complexity and processing time increase
Solution Approach 1:
The patent applies partial action by selectively using multiple color palettes only where needed for defect analysis, rather than applying complex processing uniformly across all image data. The system uses different color palettes specifically for comparison image display to enhance defect visibility, while keeping other processing relatively simple, thus improving detection accuracy without proportionally increasing processing time
3Measurement precision
If conventional single-color palette representation is used, then the system is easy to operate, but operators cannot easily differentiate between defects and pattern edges
Solution Approach 1:
The patent employs color changes by assigning different color palettes to different display purposes. The comparison image uses a distinct color palette from the optical image, creating visual differentiation that helps operators easily distinguish defects from pattern edges. This color differentiation maintains ease of operation by providing intuitive visual cues without requiring complex operational procedures
Data Source
AI summary
A review apparatus comprises: means for acquiring a reference image and an optical image by associating the data of the reference image and the data of the optical image with the gradation values defined by a color palette selected from among a plurality of first color palettes; and means for acquiring a comparison image by associating comparison image data that can be acquired from the reference image data and the optical image data with the gradation values defined by a color palette selected from among a plurality of second color palettes. Gradation values for the comparison image data are represented by gradation values for the reference image data and gradation values for the optical image data.