Abstract
The assessment of aluminum melt cleanliness has traditionally relied on labor-intensive and subjective manual processes. The present study builds upon prior digital image analysis to quantify inclusions in micrographs of PoDFA samples. Through the integration of deterministic methods, unsupervised Machine Learning (ML), and neural networks, cleanliness data comparable to PoDFA grid assessments has been achieved. Overcoming the challenge of generating sufficient and accurate training data for neural networks, the suggested approach has been refined. Enhanced isolation strategies for target classes have resulted in higher-quality training data, elevating the prediction accuracy of the neural network. Post-processing of neural network predictions has also been improved. The integrated approach presented here demonstrates more reliable cleanliness data than previous implementations. Offering a promising alternative to manual PoDFA assessments, this integrated approach improves efficiency and reduces human biases.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
“Norsk Hydro.” Accessed online: Oct. 08, 2023. Available: www.hydro.com
A. K. Nayak, H. Zedel, S. Akhtar, R. Fritzsch, and R. E. Aune, “Automated Image Analysis of Metallurgical Grade Samples Reinforced with Machine Learning,” 2023, pp. 890–897. https://doi.org/10.1007/978-3-031-22532-1_118
H. Zedel, R. Fritzsch, S. Akhtar, and R. E. Aune, “Automated Metal Cleanliness Analyzer (AMCA) – An Alternative Assessment of Metal Cleanliness in Aluminium Melts”
O. Ronneberger, P. Fischer, and T. Brox, “U-Net: Convolutional Networks for Biomedical Image Segmentation,” ArXiv
H. Zedel, R. Fritzsch, S. Akhtar, and R. E. Aune, “Automated Metal Cleanliness Analyzer (AMCA): Digital Image Analysis Phase Differentiation and Benchmarking Against PoDFA-Derived Cleanliness Data,” 2023, pp. 882–889. https://doi.org/10.1007/978-3-031-22532-1_117
P. V. Evans, P. G. Enright, and R. A. Ricks, “Molten Metal Cleanliness: Recent Developments to Improve Measurement Reliability,” 2018, pp. 839–846. https://doi.org/10.1007/978-3-319-72284-9_109
J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, “You Only Look Once: Unified, Real-Time Object Detection,” ArXiv
“Dragonfly | ORS.” Accessed online: Oct. 08, 2023. Available: www.theobjects.com > dragonfly
“Detectron2 | Meta.” Accessed online: Oct. 08, 2023. Available: www.ai.meta.com/tools/detectron2/
K. Simonyan and A. Zisserman, “Very Deep Convolutional Networks for Large-Scale Image Recognition,” ArXiv
Acknowledgements
The authors wish to express their gratitude to the Norwegian University of Science and Technology (NTNU), Department of Materials Science and Engineering, as well as the Department of Chemistry, for their continuous support. Special thanks are also conveyed to Norsk Hydro ASA in Sunndalsøra and Karmøy, Norway, for generously providing micrographs of PoDFA samples and sustaining continued support for the project.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2024 The Minerals, Metals & Materials Society
About this paper
Cite this paper
Nayak, A.K., Zedel, H., Akhtar, S., Fritzsch, R., Aune, R.E. (2024). Enhancing Quantification of Inclusions in PoDFA Micrographs Through Integration of Deterministic and Deep Learning Image Analysis Algorithms. In: Wagstaff, S. (eds) Light Metals 2024. TMS 2024. The Minerals, Metals & Materials Series. Springer, Cham. https://doi.org/10.1007/978-3-031-50308-5_124
Download citation
DOI: https://doi.org/10.1007/978-3-031-50308-5_124
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-50307-8
Online ISBN: 978-3-031-50308-5
eBook Packages: Chemistry and Materials ScienceChemistry and Material Science (R0)