|About this Abstract
||2019 TMS Annual Meeting & Exhibition
||Computational Approaches for Big Data, Artificial Intelligence and Uncertainty Quantification in Computational Materials Science
||Steel Inclusion Classification Using Computer Vision and Machine Learning
||Nan Gao, Mohammad Abdulsalam, Bryan Webler, Elizabeth Holm
|On-Site Speaker (Planned)
Inclusions form during liquid steel processing and affect subsequent material performance. However, it is hard to classify inclusion composition visually via scanning electron microscope (SEM) images. Energy dispersive x-ray spectroscopy (EDS) is therefore used for inclusion classification, but it adds to analysis time. In our work, we use computer vision and machine learning techniques to automatically classify steel inclusions in a large database of SEM images. We explore classifying these particles on the basis of image intensity distributions using different pre-processing and analysis methods. This approach is motivated by the difference in contrast in well-calibrated SEM images that arises from differences in chemical composition. The overall objective of this work is to automatically classify the different types of inclusion in steel directly from SEM images without the need for EDS.
||Planned: Supplemental Proceedings volume