Inverse Linescan Edge Detection for Unbiased Roughness Measurement

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods for measuring roughness of pattern structures in noisy images, such as those from scanning electron microscopes, face challenges in differentiating noise from actual roughness, leading to biased measurements that are sensitive to metrology tool settings and prone to errors due to stochastic variations at the molecular scale.

Innovation Solution

A system and method that use an inverse linescan model to detect edges in pattern structures without filtering, generating a biased power spectral density dataset, evaluating noise models, and subtracting noise to obtain unbiased roughness measures, while filtering out artifacts to reveal true PSD behavior.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If noise filtering is applied to images before roughness measurement, then measurement reliability is improved, but measurement precision deteriorates due to loss of high-frequency roughness information

Engineering Contradiction:

Improvemeasurement reliabilityVSAvoidroughness measurement precision

Core Design Contradiction:

ReliabilityVSMeasurement precision

Solution Approach 1:

The PSD dataset is segmented into multiple frequency portions (low-frequency, mid-frequency, high-frequency ranges). This segmentation allows selective processing where noise is removed from certain frequency ranges while preserving roughness information in others, resolving the contradiction between noise removal and roughness preservation

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Different processing approaches are applied to different frequency portions of the PSD dataset. The noise model is evaluated and applied selectively to specific frequency ranges where noise dominates, while preserving the high-frequency portions that contain actual roughness information, thus maintaining both reliability and precision

Inventive Principle:

Principle #3Local quality

2Ease of operation

If conventional roughness measurement methods are used on noisy images, then ease of operation is maintained, but measurement precision deteriorates due to noise contamination

Engineering Contradiction:

Improvemeasurement easeVSAvoidroughness measurement precision

Core Design Contradiction:

Ease of operationVSMeasurement precision

Solution Approach 1:

A noise model is evaluated from the PSD dataset before final roughness measurement. This preliminary noise characterization allows the noise to be subtracted or corrected in subsequent processing steps, improving measurement precision while maintaining ease of operation through automated model-based correction

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The PSD dataset serves as an intermediary representation between the raw noisy image and the final roughness measurement. By transforming the problem into the frequency domain through PSD analysis, noise can be selectively removed while preserving roughness characteristics, then transforming back to obtain accurate roughness measurements

Inventive Principle:

Principle #24Intermediary (Mediator)

3Reliability

If metrology tool settings are adjusted to reduce noise, then measurement reliability is improved, but device complexity increases due to sensitivity to multiple settings

Engineering Contradiction:

Improvemeasurement reliabilityVSAvoidmetrology tool settings complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The approach transforms the measurement parameters from spatial domain image properties to frequency domain PSD characteristics. By evaluating the PSD dataset and identifying noise characteristics in the frequency domain, the method reduces sensitivity to metrology tool settings while maintaining reliability through parameter-based noise correction

Inventive Principle:

Principle #35Parameter changes

4Measurement precision

If images are processed without filtering to preserve roughness information, then measurement precision is improved, but reliability deteriorates due to noise contamination

Engineering Contradiction:

Improveroughness measurement precisionVSAvoidmeasurement reliability

Core Design Contradiction:

Measurement precisionVSReliability

Solution Approach 1:

The noise present in the unfiltered images is converted into useful information through PSD analysis. By evaluating the high-frequency portion of the PSD dataset, the noise characteristics are identified and quantified, then this noise model is used to correct the roughness measurements, transforming the harmful noise into a beneficial correction mechanism

Inventive Principle:

Principle #22Blessing in disguise (Convert harm into benefit)

Data Source

PatentUS20230326711A1System and method for generating and analyzing roughness measurements

Publication Date: 2023.10.12 FRACTILIA LLC

AI summary

In one embodiment, a method for detecting edge positions in a pattern structure is disclosed. The method includes detecting the edge positions of features within the pattern structure of an image without filtering the image, wherein the detecting is performed by: applying a model to a single linescan, adjusting, based on the single linescan, one or more parameters of the model, obtaining, using the one or more adjusted parameters, a best fit of the model to the single linescan.