Microlithographic Image Analysis via Edge Fragment Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The analysis of microlithographic microstructured components, such as masks and wafers, faces challenges in reliably distinguishing between coated and uncoated regions in highly pixellated and noisy images, especially when regions have similar brightness, leading to inaccurate contour extraction and segmentation.
Innovation Solution
A method that isolates edge fragments, classifies them as relevant or irrelevant based on average intensity gradients, and ascertains contiguous segments without prior gap closure, eliminating irrelevant fragments and branches to enhance accuracy and reduce computation time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If conventional contour extraction methods using second derivative and filter mechanisms are applied to strongly pixellated and noisy images, then the processing can be performed with standard algorithms, but the reliability of distinguishing between coated and uncoated regions deteriorates when regions have similar brightness
Solution Approach 1:
The patent segments the edge detection process into multiple stages: initial edge fragment isolation, classification into relevant/irrelevant fragments, and final contour construction. This multi-stage segmentation allows the system to handle noisy, pixellated images more reliably by processing edge information in discrete, manageable steps rather than applying a single conventional filter.
Solution Approach 2:
The patent performs preliminary classification of edge fragments as relevant or irrelevant before final contour extraction. By pre-processing and categorizing edge fragments based on local intensity gradients and geometric properties, the system prepares the data in advance to improve subsequent contour detection accuracy, especially in regions with similar brightness where conventional methods fail.
2Productivity
If threshold-based contour detection using intensity values is applied, then the method is computationally simple, but the accuracy deteriorates in highly pixellated images where coated and uncoated regions have similar average brightness
Solution Approach 1:
Instead of applying a global threshold to the entire image, the patent applies local quality assessment by evaluating intensity gradients and edge fragment properties at each local position. This allows the system to adapt to local variations in brightness and noise characteristics, maintaining high measurement precision even when coated and uncoated regions have similar average brightness values.
Solution Approach 2:
The patent changes the detection parameter from absolute intensity values to intensity gradients and edge fragment characteristics. By transforming the detection basis from raw intensity (zero-order derivative) to gradient-based metrics (first-order derivative and beyond), the system achieves better discrimination between regions with similar brightness while maintaining computational efficiency through localized processing.
3Reliability
If gap closure operations are performed before contour extraction to handle interrupted edges, then contour continuity is improved, but computation time and processing complexity increase significantly
Solution Approach 1:
The patent performs preliminary classification and filtering of edge fragments before final contour extraction, preparing the data structure in advance to enable direct contour construction without time-consuming gap closure operations. By pre-organizing relevant edge fragments and their spatial relationships, the system achieves contour continuity efficiently during the extraction phase rather than through separate, computationally intensive closing operations.
Solution Approach 2:
The patent extracts and isolates relevant edge fragments from the noisy image data before contour construction. By separating and categorizing only the necessary edge information (relevant fragments) from irrelevant data, the system reduces the computational burden of subsequent contour extraction and eliminates the need for extensive gap closure operations on the entire image, thereby reducing computation time while maintaining contour continuity.
Data Source
AI summary
The invention relates to a method and to an apparatus for analyzing an image of a microlithographic microstructured component wherein in the image each of a multiplicity of pixels is assigned in each case an intensity value. A method according to the invention comprises the following steps: isolating a plurality of edge fragments in the image;classifying each of the isolated edge fragments either as a relevant edge fragment or as an irrelevant edge fragment; and ascertaining contiguous segments in the image based on the relevant edge fragments.