Deep-layers-assisted machine learning for accurate image segmentation of complex materials
Source abstract
Abstract
In materials science, X-ray micro-computed tomography (μCT) imagery provides a unique tool for quantitative and qualitative characterization of structure and performance. The tomographic analysis relies on the segmentation of large data sets to produce quantifiable volumes, a highly time-intensive process. Great progress has been observed in developing robust and efficient segmentation techniques, and machine learning currently stands out as a powerful segmentation tool. These methods have demonstrated excellent performance when applied to various materials. However, the performance of such methods heavily relies on the amount of ground truth data. This presents a great challenge in image segmentation since the amount of a priori raw images can be limited by sample cost and size. Additionally, manual labeling of data sets is often very time-intensive and can be subjected to human error. Therefore, a machine learning method is needed which achieves high segmentation accuracies for reduced training data sets. To this end, we have offered a stochastic method for producing variations of the training images that will retain the important class-wide features and thereby enrich the machine learning`s “understanding” of the variabilities. The proposed method can significantly increase the final overall accuracy. We found that by enlarging the initial training set by additional realizations, we were able to improve the average accuracy of segmentation from 81.1% to 90.0% for a very complex Mg-based alloy. The results of this study show that it is possible to increase the accuracy of predictions in imaging from X-ray microscopy using machine learning methods when enough data are not available.
Source: expertdata · Quality: candidate unverified · Source record ↗
Research context
OpenAlex topics: Nuclear Physics and Applications · Machine Learning in Materials Science · Advanced X-ray and CT Imaging
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (91)
Show source-provided references
1. Quantitative X-ray tomography10.1179/1743280413Y.0000000023 ↗
2. Development of a lab-scale, high-resolution, tube-generated X-ray computed-tomography system for three-dimensional (3D) materials characterization10.1016/j.matchar.2014.03.002 ↗
3. Crack closure and stress intensity factor measurements in nodular graphite cast iron using three-dimensional correlation of laboratory X-ray microtomography images10.1016/j.actamat.2009.05.005 ↗
4. The ascent of 3D X-ray microscopy in the laboratory10.1017/S1551929513000060 ↗
5. Advances in synchrotron radiation microtomography10.1016/j.scriptamat.2006.02.012 ↗
6. X-ray microscopy with synchrotron radiation10.1364/AO.15.001883 ↗
7. Damage evolution in SiC particle reinforced Al alloy matrix composites by X-ray synchrotron tomography10.1016/j.actamat.2010.07.039 ↗
8. Three-dimensional (3D) microstructure-based modeling of a thermally-aged cast duplex stainless steel based on X-ray microtomography, nanoindentation and micropillar compression
9. Formation of intermetallic δ phase in Al-10Si-0.3Fe alloy investigated by in-situ 4D X-ray synchrotron tomography10.1016/j.actamat.2017.02.048 ↗
10. Upscaling ice crystal growth dynamics in snow: rigorous modeling and comparison to 4D X-ray tomography data10.1016/j.actamat.2018.03.010 ↗
11. In situ and real-time 3-D microtomography investigation of dendritic solidification in an Al-10 wt% Cu alloy10.1016/j.actamat.2009.01.035 ↗
12. Probing novel microstructural evolution mechanisms in aluminum alloys using 4D nanoscale characterization10.1002/adma.201703482 ↗
13. Following the phase transitions of iron in 3D with X-ray tomography and diffraction under extreme conditions10.1016/j.actamat.2020.04.030 ↗
14. Multiscale investigation of corrosion damage initiation and propagation in AA7075-T651 alloy using correlative microscopy10.1016/j.corsci.2021.109429 ↗
15. Electromigration in Bi-crystal pure Sn solder joints: elucidating the role of grain orientation10.1016/j.jallcom.2019.152918 ↗
16. 3D time-resolved observations of fatigue crack initiation and growth from corrosion pits in Al 7XXX alloys using in situ synchrotron X-ray tomography10.1007/s11661-019-05519-z ↗
17. Four dimensional (4D) microstructural evolution of Cu6Sn5 intermetallic and voids under electromigration in bi-crystal pure Sn solder joints10.1016/j.actamat.2020.02.052 ↗
18. Damage investigation and modeling of 3D woven ceramic matrix composites from X-ray tomography in-situ tensile tests10.1016/j.actamat.2017.08.034 ↗
19. In situ X-ray microtomography of stress corrosion cracking and corrosion fatigue in aluminum alloys10.1007/s11837-017-2413-8 ↗
20. Microstructural design for damage tolerance in high strength steels10.1016/j.matlet.2020.127664 ↗
21. In situ 3D quantification of the evolution of creep cavity size, shape, and spatial orientation using synchrotron X-ray tomography10.1016/j.msea.2007.05.108 ↗
22. Direct observation of the displacement field and microcracking in a glass by means of X-ray tomography during in situ Vickers indentation experiment10.1016/j.actamat.2019.07.053 ↗
23. Real-time quantitative imaging of failure events in materials under load at temperatures above 1600 °C10.1038/nmat3497 ↗
24. In situ X-ray microtomography of the compression behavior of eTPU bead foams with a unique graded structure
25. Synchrotron CT imaging of lattice structures with engineered defects10.1007/s10853-020-04840-y ↗
26. In situ and real-time 3-D microtomography investigation of dendritic solidification in an Al-10 wt% Cu alloy10.1016/j.actamat.2009.01.035 ↗
27. Real-time quantitative imaging of failure events in materials under load at temperatures above 1600 °C10.1038/nmat3497 ↗
28. Probing novel microstructural evolution mechanisms in aluminum alloys using 4D nanoscale characterization10.1002/adma.201703482 ↗
29. Beam hardening artifacts in micro-computed tomography scanning can be reduced by X-ray beam filtration and the resulting images can be used to accurately measure BMD10.1016/j.bone.2009.07.078 ↗
30. A method for zinger artifact reduction in high-energy x-ray computed tomography
31. Interactive segmentation with intelligent scissors10.1006/gmip.1998.0480 ↗
32. The investigation of morphometric parameters of aluminum foams using micro-computed tomography10.1016/S0921-5093(01)01809-3 ↗
33. Characterization of the morphology of cellular ceramics by 3D image processing of X-ray tomography10.1016/j.jeurceramsoc.2006.05.097 ↗
34. State of 3-D micro-damage in hydrogen redistributed regions of precharged high strength aluminum alloy10.1016/j.corsci.2016.04.050 ↗
35. Quantitative assessment of microstructure and its effects on compression behavior of aluminum foams via high-resolution synchrotron x-ray tomography10.1007/s11661-006-1072-0 ↗
36. A 3D measurement procedure for internal local crack driving forces via synchrotron X-ray microtomography10.1016/j.actamat.2003.11.014 ↗
37. Direct measurement procedure for three-dimensional local crack driving force using synchrotron X-ray microtomography10.1016/j.actamat.2008.08.022 ↗
38. Segmentation of X-ray computed tomography images of porous materials: a crucial step for characterization and quantitative analysis of pore structures10.1029/2009WR008087 ↗
39. Automatic measurement of sister chromatid exchange frequency10.1177/25.7.70454 ↗
40. Unimodal thresholding10.1016/S0031-3203(00)00136-9 ↗
41. An algorithm for fast adaptive image binarization with applications in radiotherapy imaging10.1109/TMI.2002.806431 ↗
42. On minimum cross-entropy thresholding10.1016/0031-3203(95)00111-5 ↗
43. Entropic thresholding10.1016/0165-1684(89)90090-X ↗
44. Image thresholding by indicator kriging
45. Image segmentation using probabilistic fuzzy c-means clustering
46. Comparison of image segmentation methods in simulated 2D and 3D microtomographic images of soil aggregates10.1016/j.geoderma.2011.01.006 ↗
47. Bayesian image classification using Markov random fields10.1016/0262-8856(95)01072-6 ↗
48. Defect detection in textured materials using optimized filters
49. Development and experimental validation of a texture-based 3D liberation model10.1016/j.mineng.2021.106828 ↗
50. Bishop · 2022
51. LeCun · 1998
52. ImageNet large scale visual recognition challenge10.1007/s11263-015-0816-y ↗
53. Deep learning10.1038/nature14539 ↗
54. Deep learning earth observation classification using ImageNet pretrained networks10.1109/LGRS.2015.2499239 ↗
55. Automated correlative segmentation of large Transmission X-ray Microscopy (TXM) tomograms using deep learning10.1016/j.matchar.2018.05.053 ↗
56. Deep learning-based image segmentation for AlLa alloy microscopic images10.3390/sym10040107 ↗
57. Characterization of metals in four dimensions10.1080/21663831.2020.1809544 ↗
58. Advanced deep learning-based 3D microstructural characterization of multiphase metal matrix composites10.1002/adem.201901197 ↗
59. Microstructure instance segmentation from aluminum alloy metallographic image using different loss functions
60. A convolutional neural network approach to calibrating the rotation axis for X-ray computed tomography10.1107/S1600577516020117 ↗
61. Low-dose x-ray tomography through a deep convolutional neural network
62. Machine-Learning-based Algorithms for Automated Image Segmentation Techniques of Transmission X-ray Microscopy (TXM)10.1007/s11837-021-04706-x ↗
63. High-fidelity geometry generation from CT data using convolutional neural networks
64. Automated correlative segmentation of large Transmission X-ray Microscopy (TXM) tomograms using deep learning10.1016/j.matchar.2018.05.053 ↗
65. Deep learning-based image segmentation for AlLa alloy microscopic images10.3390/sym10040107 ↗
66. Advanced deep learning-based 3D microstructural characterization of multiphase metal matrix composites10.1002/adem.201901197 ↗
67. Machine-learning-based algorithms for automated image segmentation techniques of transmission X-ray microscopy (TXM)10.1007/s11837-021-04706-x ↗
68. Transferring deep convolutional neural networks for the scene classification of high-resolution remote sensing imagery
69. Enhancing images of shale formations by a hybrid stochastic and deep learning algorithm10.1016/j.neunet.2019.07.009 ↗
70. Segmentation of digital rock images using deep convolutional autoencoder networks10.1016/j.cageo.2019.02.003 ↗
71. Reconstruction of three-dimensional porous media using a single thin section10.1103/PhysRevE.85.066709 ↗
72. Cross-correlation function for accurate reconstruction of heterogeneous media10.1103/PhysRevLett.110.078002 ↗
73. Enhancing multiple-point geostatistical modeling: 1. Graph theory and pattern adjustment
74. Enhancing multiple-point geostatistical modeling: 2. Iterative simulation and multiple distance function
75. Enhancing images of shale formations by a hybrid stochastic and deep learning algorithm10.1016/j.neunet.2019.07.009 ↗
76. Accurate modeling and evaluation of microstructures in complex materials10.1103/PhysRevE.97.023307 ↗
77. Packing of discrete and irregular particles10.1016/j.compgeo.2018.03.011 ↗
78. Adam: a method for stochastic optimization
79. U-net: Convolutional networks for biomedical image segmentation
80. A survey on deep learning techniques for image and video semantic segmentation10.1016/j.asoc.2018.05.018 ↗
81. Learning deconvolution network for semantic segmentation
82. Nucleation and growth crystallography of Al8Mn5 on B2-Al(Mn,Fe) in AZ91 magnesium alloys10.1016/j.actamat.2018.04.032 ↗
83. Braszczy · 2011 · 95
84. Characterization of AlMn particles in AZ91D investment castings10.1016/j.matchar.2006.10.010 ↗
85. High-throughput x-ray microtomography at the Advanced Photon Source beamline 2-BM
86. Synchrotron Radiation TomoPy: a framework for the analysis of synchrotron tomographic data10.1107/S1600577514013939 ↗
87. Developments in synchrotron x-ray computed microtomography at the National Synchrotron Light Source
88. Simultaneous phase and amplitude extraction from a single defocused image of a homogeneous object10.1046/j.1365-2818.2002.01010.x ↗
89. Reconstruction formulas in phase-contrast tomography10.1016/S0030-4018(99)00575-1 ↗
90. Stripe and ring artifact removal with combined wavelet-Fourier filtering
91. Keras