SEM Coating Thickness Analysis for Overlapping Grayscale Phases

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing thresholding methods fail to accurately segment and measure the thickness of boron nitride (BN) and silicon-doped boron nitride (SiBN) layers in next-generation interfacial coatings (IFCs) due to overlapping grayscale intensities, leading to errors in mechanical property evaluation of fiber-matrix interfaces in ceramic matrix composites.

Innovation Solution

A method and system using scanning electron microscopy (SEM) with backscattered electron detection, employing image processing software to identify inflection points in grayscale gradients for phase segmentation and thickness measurement, utilizing a maximum inscribed circle method and watershed algorithm to automate the analysis of SEM images.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If traditional thresholding techniques are used for phase segmentation, then the method is simple and easy to implement, but the measurement precision deteriorates when grayscale intensities of different phases overlap

Engineering Contradiction:
Improveease of implementationVSAvoidthickness measurement precision
Core Design Contradiction:
Ease of manufactureVSMeasurement precision

Solution Approach 1:

The patent transforms the segmentation approach from direct grayscale thresholding to gradient-based inflection point detection. By calculating the gradient of grayscale values and identifying inflection points where the gradient reaches local maxima, the method adapts to varying grayscale intensities of different phases (BN, SiBN, SiC) without requiring manual threshold adjustment, thereby maintaining both ease of implementation and measurement precision

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent transitions from analyzing grayscale intensity values in one dimension to analyzing the gradient of grayscale values in a transformed dimension. This dimensional transformation allows the method to detect phase boundaries through inflection points in the gradient profile, which remain distinct even when the original grayscale intensities of adjacent phases overlap

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Adaptability or versatility

If standard thresholding methods are applied to next-generation IFCs with SiBN layers, then the analysis routine remains consistent with traditional methods, but the reliability of segmentation deteriorates due to overlapping intensities

Engineering Contradiction:
Improvemethod consistencyVSAvoidsegmentation reliability
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The patent modifies the analysis parameter from direct grayscale thresholding to gradient-based inflection point detection. This parameter change enables the method to reliably distinguish between BN, SiBN, and SiC phases by detecting the characteristic inflection points in the gradient profile, which remain identifiable even when grayscale intensities overlap, thereby maintaining segmentation reliability across different IFC compositions

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent introduces the gradient of grayscale values as an intermediary parameter between the original grayscale image and the final segmentation result. This intermediary transformation highlights phase boundaries through inflection points, serving as a reliable mediator that connects the input image data to accurate thickness measurements even when direct thresholding would fail

Inventive Principle:
Principle #24Intermediary (Mediator)

3Measurement precision

If manual threshold adjustment is performed for each image or batch, then the measurement precision can be improved, but the productivity deteriorates due to time-consuming manual intervention

Engineering Contradiction:
Improvethickness measurement precisionVSAvoidanalysis throughput
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent implements an automated inflection point detection algorithm that self-adjusts to identify phase boundaries without manual intervention. The algorithm automatically calculates gradients, detects inflection points, and determines thickness measurements, enabling the system to process multiple images and batches efficiently while maintaining high measurement precision through adaptive boundary detection

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent employs a feedback mechanism where the gradient calculation and inflection point detection automatically adapt to the specific grayscale characteristics of each image. The algorithm uses the detected inflection points to refine boundary identification, creating a self-correcting system that maintains precision across varying conditions without requiring manual threshold adjustment for each case

Inventive Principle:
Principle #23Feedback

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

Accurately segments and measures BN and SiBN thicknesses in IFCs, enabling reliable mechanical property evaluation of fiber-matrix interfaces, applicable to ceramic matrix composites and other multiphase materials with similar grayscale contrasts.

Implementation Method 1

Scanning electron microscopy (SEM) imaging with a backscattered electron detector (BSD) can be used to produce images of the material layers

Methodology Applied
Scientific EffectBackscattered electron detection:

Data Source

PatentEP4495542B1Image analysis method for multi-phase system with overlapping grayscale intensities
Publication Date: 2026.04.08 RTX CORP
  • EP4495542B1 patent drawingFigure 1
  • EP4495542B1 patent drawingFigure 2
  • EP4495542B1 patent drawingFigure 3

AI summary

A method for determining coating thicknesses of a coated fiber embedded in a matrix is presented. A scanning electron microscope (SEM) image is captured of a composite material having multiple coated fibers embedded in a matrix. This image depicts a cross-sectional portion of the composite material. A processor identifies a region of the image depicting a cross-section of one among the plurality of coated fibers, and generates a line graph of a gradient of grayscale values as a function of a line scan. The line scan traverses across the image along a radius of the coated fiber, and extends between an interior location of the coated fiber and an exterior region outside the coated fiber. The processor identifies inflection points in the line graph of the gradient of grayscale values as a function of the line scan traverse location, and determines coating thickness based on these inflection points.