Level Set Method for High-Resolution Lithographic Image Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting and quantifying device structures in lithographic images using SEMs result in low-resolution representations due to noise reduction techniques, leading to loss of information and simplistic representations of high-resolution structures.
Innovation Solution
The application of level set methods and curve evolution techniques to create mathematical representations of device structures, allowing for high-resolution detection and registration of images while preserving the general shape and details of the structures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If smoothing, adaptive thresholding, edge-detection, erosion, and dilation techniques are used to process noisy images, then noise is reduced and edge detection is automated, but image resolution is lost and information is lost
Solution Approach 1:
The patent replaces traditional mechanical image processing operations (smoothing, thresholding, erosion, dilation) with a level set method that uses curve evolution and numerical functions. This substitution allows the system to detect edges and represent structures without the resolution-loss inherent in conventional mechanical processing steps.
Solution Approach 2:
The patent changes the fundamental parameter representation from discrete pixel values to continuous numerical functions (level set functions). By evolving curves based on image gradients and representing structures through these continuous functions, the system maintains high-resolution information while achieving noise reduction and edge detection.
2Extent of automation
If traditional contour models using image gradients are used for edge detection, then automated edge detection is achieved, but the quantification becomes low-resolution and simplistic
Solution Approach 1:
The patent replaces traditional contour models with a level set framework that uses evolving curves and numerical functions. This substitution maintains automation while achieving high-resolution quantification by representing structures through continuous mathematical functions rather than discrete pixel-based contours.
Solution Approach 2:
The patent transitions from two-dimensional pixel-based contour representation to a higher-dimensional numerical function representation. The level set function φ(x,y) exists in a higher-dimensional space, allowing it to capture fine structural details and provide high-resolution quantification while remaining automated.
3Productivity
If mathematical manipulation of images is performed to reduce noise and automate edge detection, then processing efficiency is improved, but resolution of the image is lost
Solution Approach 1:
The patent replaces sequential mathematical manipulation steps (smoothing, thresholding, edge detection) with a unified level set method. This single framework performs all functions simultaneously through curve evolution, maintaining processing efficiency while preserving resolution by avoiding multiple resampling operations.
Solution Approach 2:
The patent maintains continuous useful action by evolving curves continuously through the image domain using the level set function. This continuous evolution process preserves resolution information throughout the processing, unlike discrete manipulation steps that repeatedly sample and lose information.
Data Source
AI summary
A method and apparatus of detection, registration and quantification of an image is described. The method may include obtaining an image of a lithographically created structure, and applying a level set method to an object, representing the structure, of the image to create a mathematical representation of the structure. The method may include obtaining a first dataset representative of a reference image object of a structure at a nominal condition of a parameter, and obtaining second dataset representative of a template image object of the structure at a non-nominal condition of the parameter. The method may further include obtaining a deformation field representative of changes between the first dataset and the second dataset. The deformation field may be generated by transforming the second dataset to project the template image object onto the reference image object. A dependence relationship between the deformation field and change in the parameter may be obtained.