Pattern Inspection Image Correction via Predictive Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current pattern defect inspection methods for ultrafine circuit patterns in semiconductor and LCD manufacturing face challenges in accurately detecting small defects buried in positional offsets, image expansion/shrink, and sensing noises, leading to image degradation and complexity in setting correction parameters.
Innovation Solution
The integration of two-dimensional input/output predictive modeling for image correction, which designates specific correction regions, generates simultaneous equations, and solves for model parameters to create a corrected pattern image, reducing image degradation and parameter complexity while focusing on sensitive inspection areas.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple sequential correction processes (alignment, image expansion/shrink correction, image swell correction, noise averaging) are performed to improve inspection accuracy, then measurement precision is improved, but device complexity and parameter setup difficulty increase
Solution Approach 1:
The patent combines multiple separate correction processes (alignment correction, image expansion/shrink correction, image swell correction, noise averaging) into a single integrated correction operation. By merging these corrections into one unified process executed simultaneously, the system achieves the same inspection accuracy improvement without the accumulated errors and complexity of sequential processing.
Solution Approach 2:
The correction unit is designed to perform multiple types of corrections (alignment, expansion/shrink, swell, noise filtering) through a single universal correction mechanism. This multi-functional approach allows one correction process to handle all correction needs that previously required separate dedicated processes, reducing overall system complexity.
2Measurement precision
If multiple sequential correction processes are performed to improve inspection accuracy, then measurement precision is improved, but loss of time increases due to accumulated errors and re-corrections
Solution Approach 1:
By merging all correction operations into a single simultaneous process, the system eliminates the time required for multiple sequential corrections and the additional time needed to correct accumulated errors from repeated operations. The integrated correction achieves the same precision in one pass rather than through multiple time-consuming stages.
3Measurement precision
If a great number of parameters are set for respective corrections to achieve high accuracy, then measurement precision is improved, but ease of operation decreases
Solution Approach 1:
The correction unit is designed as a universal correction mechanism that handles multiple correction types (alignment, expansion/shrink, swell, noise filtering) through a unified parameter set. This eliminates the need for operators to configure separate parameters for each correction process, significantly improving ease of operation while maintaining high inspection accuracy.
4Measurement precision
If repeated execution of correction processes is performed to improve accuracy, then measurement precision is improved, but object-generated harmful factors increase due to image degradation and corruption
Solution Approach 1:
By combining all correction operations into a single simultaneous process rather than repeating corrections multiple times, the system achieves the desired accuracy improvement without subjecting the image to repeated correction operations. This prevents the accumulation of correction-induced artifacts and degradation that occur with repeated processing.
Data Source
AI summary
A high-accuracy image correction device adaptable for use in pattern inspection apparatus is disclosed. The device includes a correction region designation unit which designates a correction region including a pattern and its nearby portion within each of an inspection reference pattern image and a pattern image under test. The device also includes an equation generator which generates by linear predictive modeling a set of simultaneous equations for a reference pattern image within the correction region and an under-test pattern image within the correction region, a parameter generator for solving the equations to obtain more than one model parameter, and a corrected pattern image generator for using the model parameter to apply the linear predictive modeling to the reference pattern image to thereby generate a corrected pattern image. A pattern inspection method using the image correction technique is also disclosed.