journal article·2021-10

A deep learning method for extensible microstructural quantification of DP steel enhanced by physical metallurgy-guided data augmentation

Chunguang Shen, Xiaolu Wei, Chenchong Wang, Wei Xu

Publisher page

Abstract

Abstract The small sample problem caused by time-consuming annotation has greatly restricted the development of deep learning (DL) methods with high generality and extensibility, hindering the wide application of DL-based microstructural analysis. To address this problem, an extensible microstructural quantification method is proposed by combining physical metallurgy (PM)-guided data augmentation and DL. In this method, PM-guided data augmentation, which combines key PM information highly related to microstructural evaluation and random image transformation, is used to generate microstructures with other processes and therefore construct a comprehensive dataset. The proposed method is successfully applied and validated to dual-phase (DP) steels to classify martensite/ferrite phases and quantify their fractions and grain sizes at various temperatures based on only two SEM images of annealing temperatures of 750 and 780 °C, in which a comprehensive dataset covering the entire dual phase regime is constructed with guidance from the content and morphology of the phase and dissolved carbide during annealing. Moreover, the good extensibility and data independency of the present method is demonstrated by various DP steels with different compositions and processes collected from the literature.
Source: openaire · Quality: candidate unverified · Source record ↗

Research context

OpenAlex topics: Microstructure and Mechanical Properties of Steels · Hydrogen embrittlement and corrosion behaviors in metals · Welding Techniques and Residual Stresses

SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.

References (42)

Show source-provided references
1. Application of chord length distributions and principal component analysis for quantification and representation of diverse polycrystalline microstructures10.1016/j.matchar.2018.09.020 ↗
2. Quantitative metallographic characterization of welding microstructures in Ti-containing TWIP steel by means of image processing analysis10.1016/j.matchar.2018.10.012 ↗
3. Deep learning-assisted microstructural analysis of Ni/YSZ anode composites for solid oxide fuel cells10.1016/j.matchar.2021.110906 ↗
4. Quantification of complex-phase steel microstructure by using combined EBSD and EPMA measurements10.1016/j.matchar.2018.05.038 ↗
5. Quantification of the laves phase in advanced 9–12% Cr steels using a standard SEM10.1016/j.matchar.2004.02.003 ↗
6. Characterization of bainitic/martensitic structures formed in isothermal treatments below the Ms temperature10.1016/j.matchar.2017.04.007 ↗
7. FE-SEM in situ observation of damage evolution in tension-compression fatigue of micro-sized single-crystal copper10.1016/j.msea.2019.138218 ↗
8. Improving microstructural quantification in FIB/SEM nanotomography10.1016/j.ultramic.2017.07.017 ↗
9. Phases quantification in DP600 steel welded by GTAW process using SEM and atomic force microscopy10.1590/1980-5373-mr-2016-0172 ↗
10. Characterisation of fine-scale microstructures by electron backscatter diffraction (EBSD)10.1016/j.scriptamat.2004.05.016 ↗
11. EBSD orientation analysis based on experimental Kikuchi reference patterns10.1016/j.actamat.2020.01.053 ↗
12. Analysis of different acicular ferrite microstructures in low-carbon steels by electron backscattered diffraction. Study of their toughness behavior10.1007/s11661-003-0010-7 ↗
13. Application of EBSD to analysis of microstructures in commercial steels10.1179/174328406X130948 ↗
14. An automated method of quantifying ferrite microstructures using electron backscatter diffraction (EBSD) data10.1016/j.ultramic.2013.11.003 ↗
15. A fast instance segmentation with one-stage multi-task deep neural network for autonomous driving10.1016/j.compeleceng.2021.107194 ↗
16. Reporting on deep learning algorithms in health care10.1016/S2589-7500(19)30132-3 ↗
17. Vision-based traffic sign detection and analysis for intelligent driver assistance systems: perspectives and survey10.1109/TITS.2012.2209421 ↗
18. Advanced steel microstructural classification by deep learning methods10.1038/s41598-018-20037-5 ↗
19. Development of high accuracy segmentation model for microstructure of steel by deep learning10.2355/isijinternational.ISIJINT-2019-568 ↗
20. Deep learning-based image segmentation for Al-La alloy microscopic images10.3390/sym10040107 ↗
21. High throughput quantitative metallography for complex microstructures using deep learning: a case study in ultrahigh carbon steel10.1017/S1431927618015635 ↗
22. Deep adversarial data augmentation for extremely low data regimes10.1109/TCSVT.2020.2967419 ↗
23. Resampling and data augmentation for short-term PV output prediction based on an imbalanced sky images dataset using convolutional neural networks10.1016/j.solener.2021.05.095 ↗
24. Creating artificial images for radiology applications using generative adversarial networks (GANs) – a systematic review10.1016/j.acra.2019.12.024 ↗
25. Randaugment: practical automated data augmentation with a reduced search space
26. Race against the machine: can deep learning recognize microstructures as well as the trained human eye?10.1016/j.scriptamat.2020.10.026 ↗
27. Application of attitude tracking algorithm for face recognition based on OpenCV in the intelligent door lock10.1016/j.comcom.2020.02.003 ↗
28. Different applied median filter in salt and pepper noise10.1016/j.compeleceng.2018.01.019 ↗
29. MultiResUNet: rethinking the U-Net architecture for multimodal biomedical image segmentation10.1016/j.neunet.2019.08.025 ↗
30. A survey on deep learning in medical image analysis10.1016/j.media.2017.07.005 ↗
31. U-Net: convolutional networks for biomedical image segmentation
32. Effective and efficient batch normalization using a few uncorrelated data for statistics estimation
33. Dropout: a simple way to prevent neural networks from overfitting
34. A generic high-throughput microstructure classification and quantification method for regular SEM images of complex steel microstructures combining EBSD labeling and deep learning10.1016/j.jmst.2021.04.009 ↗
35. Importance of martensite spatial distribution at large volume fractions in imparting ductility in high-strength dual-phase steel10.1007/s11665-019-03890-x ↗
36. Effect of martensite morphology and volume fraction on strain hardening and fracture behavior of martensite–ferrite dual phase steel10.1016/j.msea.2015.01.006 ↗
37. Effect of heat-treatment schedule on the microstructure and mechanical properties of cold-rolled dual-phase steels10.1007/s40195-015-0305-x ↗
38. Deformation and fracture mechanisms in fine- and ultrafine-grained ferrite/martensite dual-phase steels and the effect of aging10.1016/j.actamat.2010.10.002 ↗
39. Visualizing data using t-SNE
40. Dimensionality reduction and visualisation of hyperspectral ink data using t-SNE10.1016/j.forsciint.2020.110194 ↗
41. Predicting strain-induced martensite in austenitic steels by combining physical modelling and machine learning10.1016/j.matdes.2020.109199 ↗
42. Estimation of deformation induced martensite in austenitic stainless steels10.1016/j.msea.2011.08.039 ↗