MetalDAM is a metallography dataset from additive manufacturing of steels. All images were kindly provided by ArcelorMittal engineers.
Dataset | Images | Download | Cite |
---|---|---|---|
Labeled | 42 | ↓ Download | Citation |
Unlabeled | 164 | ↓ Download | Citation |
The labeled set contains 42 grayscale images taken with a scanning electron microscope, with resolutions 1280x895 and 1024x703 (after cropping the information band at the bottom).
Label | Frequency |
---|---|
0. Matrix | 31.86% |
1. Austenite | 58.26% |
2. Martensite/Austenite | 8.96% |
3. Precipitate* | 0.24% |
4. Defect | 0.68% |
(*) The reduced size of the precipitates increased notably the effort of annotating these micro-constituents, so most of them were ignored during the annotation process.
An additional set of 164 images obtained from the same materials is available as well.
This dataset has been generated and provided by ArcelorMittal Global R&D. Special thanks to the researchers Margarita Guerrero Gesto, Rosalía Rementeria Fernández and Laura del Río Fernandez.
If you want to cite MetalDAM, please use the following citation:
@misc{metaldam,
title={MetalDAM: Metallography dataset from additive manufacturing},
author={{ArcelorMittal} and {DaSCI Andalusian Research Institute}},
url={https://github.com/ari-dasci/OD-MetalDAM}
}
A publication presenting the dataset will be added here soon.