Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Synergistic Sensitization of Pancreatic Cancer Cells by Nanosecond Pulsed Electric Fields and Cold Atmospheric Plasma via Amplifying ROS and Apoptotic Signaling.

International journal of molecular sciences·2026

Same author

Integrating LC-MS and network pharmacology to uncover the mechanism of antidepressant effects of Illigera aromatica ethanol extract in rodent models.

Fitoterapia·2026

Same author

Interaction between dynamic reinforcement learning and working memory of pigeon: A comparative modeling study.

The Journal of experimental biology·2026

Same author

Individual detachment-reintegration events in homing pigeon flocks and the dominance of directional adjustment in their kinematic features.

Journal of the Royal Society, Interface·2026

Same author

Correction to: Antibacterial activity and mechanism of rutin in UV-A light treatment against Escherichia coli O157:H7.

Archives of microbiology·2026

Same author

Comparative mass spectrometry analysis of high and low centrifugation extracellular vesicle (EV) pellets from healthy urine following Tamm-Horsfall protein removal.

Clinical proteomics·2026

Search research articles

Related Experiment Video

Updated: Sep 5, 2025

Visualization of Failure and the Associated Grain-Scale Mechanical Behavior of Granular Soils under Shear using Synchrotron X-Ray Micro-Tomography

Visualization of Failure and the Associated Grain-Scale Mechanical Behavior of Granular Soils under Shear using Synchrotron X-Ray Micro-Tomography

Published on: September 29, 2019

An Improved U-Net Image Segmentation Method and Its Application for Metallic Grain Size Statistics.

Peng Shi^1,2, Mengmeng Duan¹, Lifang Yang¹

¹National Center for Materials Service Safety, University of Science and Technology Beijing, Beijing 100083, China.

Materials (Basel, Switzerland)

|July 9, 2022

Summary

An improved U-Net model accurately segments metallographic images for grain size analysis, even with low contrast and fuzzy boundaries. This method requires only a small training set, outperforming traditional and other machine learning approaches.

Keywords:

complex image grain size image segmentation improved U-Net metallographic microstructure analysis

More Related Videos

Two Algorithms for High-throughput and Multi-parametric Quantification of Drosophila Neuromuscular Junction Morphology

Two Algorithms for High-throughput and Multi-parametric Quantification of Drosophila Neuromuscular Junction Morphology

Published on: May 3, 2017

Processing of Bulk Nanocrystalline Metals at the US Army Research Laboratory

Processing of Bulk Nanocrystalline Metals at the US Army Research Laboratory

Published on: March 7, 2018

Related Experiment Videos

Last Updated: Sep 5, 2025

Visualization of Failure and the Associated Grain-Scale Mechanical Behavior of Granular Soils under Shear using Synchrotron X-Ray Micro-Tomography

Visualization of Failure and the Associated Grain-Scale Mechanical Behavior of Granular Soils under Shear using Synchrotron X-Ray Micro-Tomography

Published on: September 29, 2019

Two Algorithms for High-throughput and Multi-parametric Quantification of Drosophila Neuromuscular Junction Morphology

Two Algorithms for High-throughput and Multi-parametric Quantification of Drosophila Neuromuscular Junction Morphology

Published on: May 3, 2017

Processing of Bulk Nanocrystalline Metals at the US Army Research Laboratory

Processing of Bulk Nanocrystalline Metals at the US Army Research Laboratory

Published on: March 7, 2018

Area of Science:

Materials Science
Image Analysis
Computational Methods

Background:

Accurate metallographic grain size measurement is crucial for material performance analysis.
Traditional image processing and existing machine learning methods struggle with low-contrast, blurry, or complex metallographic images.
Current machine learning approaches often require extensive datasets for effective training.

Purpose of the Study:

To develop an improved U-Net model for automated segmentation of complex metallographic images.
To address the limitations of existing methods in handling low contrast, fuzzy boundaries, and complex structures.
To enable accurate grain size analysis with minimal training data.

Main Methods:

An improved U-Net deep learning model was developed for image segmentation.
The model was trained on a small dataset of metallographic images.
Segmentation performance was evaluated using metrics such as Accuracy (ACC), Mean Intersection over Union (MIoU), Precision, and F1-score.
Grain size was calculated based on segmentation results following ASTM standards.

Main Results:

The improved U-Net model demonstrated significant advantages in segmenting challenging metallographic images.
The method achieved high performance metrics: ACC of 0.97, MIoU of 0.752, Precision of 0.98, and F1 of 0.96.
The model effectively handled images with low contrast, fuzzy boundaries, and complex structures.
Grain size calculation based on the model's segmentation yielded satisfactory results.

Conclusions:

The proposed improved U-Net model offers an effective solution for automated metallographic image segmentation, particularly for complex cases.
This approach significantly enhances the accuracy and efficiency of grain size analysis.
The method's ability to perform well with small training sets makes it a practical tool for materials science research.