Description of Ore Particles from X-Ray Microtomography (XMT) Images, Supported by Scanning Electron Microscope (SEM)-Based Image Analysis

Furat, Orkun; Leißner, Thomas; Ditscherlein, Ralf; Šedivý, Ondřej; Weber, Michael; Bachmann, K.‐H.; Gutzmer, Jens; Peuker, Urs A.; Schmidt, Volker

doi:10.1017/s1431927618015076

Cited by 33 publications

(23 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A promising way to make use of such characteristics is machine learning which has recently been shown to be a powerful tool for segmentation of X-ray tomography data. [52] Applying machine-learning or deeplearning techniques to segment PC data sets requires a better understanding of PC formation. This will be addressed in future research by systematic studies of model structures and comparable data sets.…”

Section: Discussionmentioning

confidence: 99%

Correlative Nano‐Computed Tomography and Focused Ion‐Beam Sectioning: A Case Study on a Co‐Base Superalloy Oxide Scale

et al. 2019

View full text Add to dashboard Cite

Herein, the capabilities of a correlative tomography approach combining laboratory nano X‐ray computed tomography, focused ion‐beam sectioning, and energy‐dispersive X‐ray spectroscopy are utilized for the characterization of multilayered oxide scales of Co‐base superalloys regarding their 3D morphology and chemical composition. The combination of complementary 3D imaging techniques allows for a precise and reliable segmentation of the pore space, oxide precipitates, and different phases of the initial material. Such information is instrumental to a microscopic understanding of oxidation in this new class of superalloys and key to improvement of oxidation resistance in the next‐generation high‐temperature materials.

show abstract

Section: Discussionmentioning

confidence: 99%

Correlative Nano‐Computed Tomography and Focused Ion‐Beam Sectioning: A Case Study on a Co‐Base Superalloy Oxide Scale

et al. 2019

View full text Add to dashboard Cite

show abstract

“…However, the watershed algorithm often fails for the considered data, since, for example, elongated particles are segmented into multiple fragments. In Furat et al (2018) a postprocessing step was described which utilizes machine learning techniques, more precisely a feedforward neural network, to eliminate oversegmentation.…”

Section: Segmentation Of Mineral Particle Systemsmentioning

confidence: 99%

“…In the present paper, we give a short overview of several applications in the field of materials science in which we successfully combined methods of statistical learningincluding random forests, feedforward and convolutional neural networks-with conventional image processing techniques for segmentation, classification and object detection tasks, see e.g., Furat et al (2018), Neumann et al (2019), and Petrich et al (2017). This shows the flexibility of the approach of combining conventional image processing with machine learning techniques, where the latter can be used either for preprocessing image data to increase the performance of conventional image processing algorithms or for postprocessing segmentations obtained by conventional means in order to improve segmentation qualities.…”

Section: Introductionmentioning

confidence: 99%

“…Then, in section 2.2, particulate systems of minerals are considered that are of interest in the mining industry. Here, a feedforward neural network is used to refine particle-wise segmentations obtained from the watershed algorithm (Furat et al, 2018). The watershed algorithm and feedforward neural networks are also combined in section 2.3.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Machine Learning Techniques for the Segmentation of Tomographic Image Data of Functional Materials

Furat¹,

Wang

Neumann³

et al. 2019

Front. Mater.

Self Cite

View full text Add to dashboard Cite

In this paper, various kinds of applications are presented, in which tomographic image data depicting microstructures of materials are semantically segmented by combining machine learning methods and conventional image processing steps. The main focus of this paper is the grain-wise segmentation of time-resolved CT data of an AlCu specimen which was obtained in between several Ostwald ripening steps. The poorly visible grain boundaries in 3D CT data were enhanced using convolutional neural networks (CNNs). The CNN architectures considered in this paper are a 2D U-Net, a multichannel 2D U-Net and a 3D U-Net where the latter was trained at a lower resolution due to memory limitations. For training the CNNs, ground truth information was derived from 3D X-ray diffraction (3DXRD) measurements. The grain boundary images enhanced by the CNNs were then segmented using a marker-based watershed algorithm with an additional postprocessing step for reducing oversegmentation. The segmentation results obtained by this procedure were quantitatively compared to ground truth information derived by the 3DXRD measurements. A quantitative comparison between segmentation results indicates that the 3D U-Net performs best among the considered U-Net architectures. Additionally, a scenario, in which "ground truth" data is only available in one time step, is considered. Therefore, a CNN was trained only with CT and 3DXRD data from the last measured time step. The trained network and the image processing steps were then applied to the entire series of CT scans. The resulting segmentations exhibited a similar quality compared to those obtained by the network which was trained with the entire series of CT scans.

show abstract

“…We propose a hybrid approach combining a watershed-based trinarization going back to ideas presented in Meyer & Beucher (1990) with a trinarization using a random forest algorithm, a tool from statistical learning. Another combination of tools from mathematical morphology with statistical learning has recently been used to improve the quality of particle-wise segmentation from 3D image data (Furat et al, 2018). In the present paper, the hybrid approach leads to a trinarization allowing for the computation of structural characteristics, which are only accessible via image analysis such as, e.g.…”

Section: Introductionmentioning

confidence: 99%

Characterization of the 3D microstructure of Ibuprofen tablets by means of synchrotron tomography

Neumann¹,

Cabiscol

Osenberg

et al. 2019

Journal of Microscopy

Self Cite

View full text Add to dashboard Cite

Summary A new methodology to segment the three‐dimensional (3D) internal structure of Ibuprofen tablets from synchrotron tomography is presented, introducing a physically coherent trinarization for greyscale images of Ibuprofen tablets consisting of three phases: microcrystalline cellulose, Ibuprofen and pores. For this purpose, a hybrid approach is developed combining a trinarization by means of statistical learning with a trinarization based on a watershed algorithm. This hybrid approach allows us to compute microstructure characteristics of tablets using methods of statistical image analysis. A comparison with experimental results shows that there is a significant amount of pores which is below the resolution limit. At the same time, results from image analysis let us conjecture that these pores constitute the great majority of the surface between pores and solid. Furthermore, we compute microstructure characteristics, which are experimentally not accessible such as local percolation probabilities and chord length distribution functions. Both characteristics are meaningful in order to quantify the influence of tablet compaction on its microstructure. The presented approach can be used to get better insight into the relationship between production parameters and microstructure characteristics based on 3D image data of Ibuprofen tablets manufactured under different conditions and elucidate key effects on the strength and solubility kinetics of the final formulation. Lay Description A typical formulation of uniaxial compacted Ibuprofen tablets consist of a mixture of an excipient (microcrystalline cellulose) with an active ingredient (a ground fraction of Ibuprofen). The final mechanical strength of the tablet as well as the release kinetics are strongly influenced by the underlying microstructure, i.e. the spatial arrangement of the microcrystalline cellulose and Ibuprofen within the tablet. In order to optimize the performance of the tablet, it is important to investigate the relationship between its microstructure and the corresponding production parameters. For this purpose, 3D imaging is a powerful tool as it allows computing microstructural properties such as the internal arrangement, interconnectivity and pore location and distribution, characteristics that cannot be computed by experimental characterization techniques. In the present study, a new algorithm for an accurate trinarization of 3D image data obtained by synchrotron tomography is presented. Trinarization means that we reconstruct microcrystalline cellulose, Ibuprofen and pores on the basis of the 3D images, where one can only observe different greyscale values, but not the different constituents themselves. For this purpose, a hybrid approach combining a trinarization by means of artificial intelligence with a trinarization based on a geometrically motivated algorithm is developed. This hybrid approach allows to compute microstructure characteristics of tablets using image analysis. A comparison with experimental results shows that there is a s...

show abstract

Description of Ore Particles from X-Ray Microtomography (XMT) Images, Supported by Scanning Electron Microscope (SEM)-Based Image Analysis

Cited by 33 publications

References 30 publications

Correlative Nano‐Computed Tomography and Focused Ion‐Beam Sectioning: A Case Study on a Co‐Base Superalloy Oxide Scale

Correlative Nano‐Computed Tomography and Focused Ion‐Beam Sectioning: A Case Study on a Co‐Base Superalloy Oxide Scale

Machine Learning Techniques for the Segmentation of Tomographic Image Data of Functional Materials

Characterization of the 3D microstructure of Ibuprofen tablets by means of synchrotron tomography

Contact Info

Product

Resources

About