Pattern Image Correction via Density-Adaptive Model Switching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current pattern inspection methods for ultrafine circuit patterns in semiconductor and LCD manufacturing face challenges in accurately detecting defects due to image degradation from repeated corrections and difficulties in setting parameters, especially in low-density patterns, leading to inaccuracies and increased errors.
Innovation Solution
An integrated approach combining 2D input/output predictive modeling and bicubic interpolation for image correction, automatically switching between methods based on pattern density, to minimize image degradation and parameter setup complexity, using an image correction device with equation generation, parameter calculation, sub-pixel interpolation, and error comparison units.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If repeated execution of image corrections (expansion/shrink correction, swell correction, resize correction) is performed, then image quality improvement is achieved, but accumulated errors occur causing appreciable image degradation
Solution Approach 1:
The patent changes the fundamental parameter of correction approach from sequential iterative corrections to a single-step predictive model-based correction. The predictive model directly computes the corrected image without repeated iterations, eliminating accumulated errors while maintaining high correction accuracy.
Solution Approach 2:
The patent replaces the mechanical/iterative correction process with a computational predictive model. Instead of repeatedly applying correction algorithms that accumulate errors, the system uses a predictive model that directly computes the corrected image, substituting iterative mechanics with a direct computational approach.
2Manufacturing precision
If multiple correction processes (alignment, expansion/shrink correction, swell correction, noise averaging) are sequentially performed, then image quality improves, but device complexity and parameter setup difficulty increase
Solution Approach 1:
The patent merges multiple separate correction processes (alignment, expansion/shrink correction, swell correction, noise averaging) into a single integrated predictive model-based correction step. This consolidation reduces device complexity while maintaining comprehensive correction functionality.
Solution Approach 2:
The predictive model serves multiple functions simultaneously: it performs alignment, expansion/shrink correction, swell correction, and noise reduction in a single operation. This multi-functionality eliminates the need for separate dedicated processes for each correction type, reducing overall system complexity.
3Manufacturing precision
If 2D linear predictive modeling is used for image correction, then correction accuracy improves, but the method fails for extra-low density patterns due to insufficient matrix information
Solution Approach 1:
The patent makes the correction method dynamic by automatically selecting between predictive modeling and bicubic interpolation based on pattern density. The system adapts its behavior according to the input pattern characteristics, switching from predictive modeling for high-density patterns to bicubic interpolation for low-density patterns, ensuring versatility across different pattern types.
Solution Approach 2:
The patent changes the correction method parameter based on pattern density conditions. For patterns with sufficient information content, predictive modeling is used; for patterns with insufficient information (extra-low density), the system switches to bicubic interpolation. This parameter change ensures the method works across the full range of pattern densities.
4Measurement precision
If bicubic interpolation-based alignment in units of subpixels is performed, then sub-pixel alignment accuracy improves, but image degradation occurs from repeated corrections
Solution Approach 1:
The patent replaces repeated bicubic interpolation operations with a single predictive model-based correction. The predictive model directly computes the corrected image without requiring multiple iterative interpolation steps, thereby achieving high sub-pixel alignment precision while avoiding the image degradation that accumulates from repeated corrections.
Data Source
AI summary
An image correction device for use in a pattern inspection apparatus is disclosed, which has automatic adaptability to variations in density of a pattern image of a workpiece being tested. The device is operable to identify a two-dimensional (2D) linear predictive model parameters from the pattern image of interest and determine the value of a total sum of these identified parameters. This value is then used to switch between a corrected pattern image due to the 2D linear prediction modeling and a corrected image that is interpolated by bicubic interpolation techniques. A pattern inspection method using the image correction technique is also disclosed.