Hyperspectral mineral mapping for underground mining
- 1Helmholtz-Zentrum Dresden-Rossendorf, Helmholtz-Institut Freiberg für Ressourcentechnologie, Division of Exploration, Freiberg, Germany (email@example.com)
- 2National Technical University of Athens, School of Mining and Metallurgical Engineering, Greece
- 3TheiaX GmbH, Freiberg, Germany
- 4University of the Witwatersrand, School of Geosciences, Johannesburg, South Africa
- 5IAV Hassan II, School of Geomatics and Surveying Engineering, Rabat, Morocco
- 6Deutsche Lithium GmbH, Freiberg, Germany
Future mining will increasingly require rapid and informed decisions to optimise ore extraction and valuation. In this context, the use of hyperspectral imaging has been proven to be effective for geological mapping in surface mining operations. The potential of hyperspectral methods in underground mining environments, however, remains underexplored due to challenges associated with illumination and surface water. Our contribution addresses this gap by evaluating different lighting setups and the effect of moisture on the spectral quality of hyperspectral data in a laboratory setup. We also compared three commercially available, visible-near infrared to shortwave infrared sensors to assess their suitability for underground hyperspectral scanning. As a demonstration, we acquired hyperspectral data from three adjacent outcrops in the visitor’s mine of Zinnwald, Germany, where rocks of a Late Variscan Sn-W-Li greisen-type deposit are exposed in representative underground mining conditions. A photogrammetric 3D digital outcrop model was used to correct for illumination effects in the data. We then estimated mineral abundance and lithium content across the mine face employing an adapted workflow that combines quantitative XRD measurements with hyperspectral unmixing techniques. Laser-induced breakdown spectroscopy was used to validate the results. While there are still challenges to overcome, this study proves that hyperspectral imaging techniques can be applied underground to yield rapid and accurate geological information. This application will pave the way for the safe, digital and automated underground mine of the future.
How to cite: Kirsch, M., Mavroudi, M., Thiele, S., Lorenz, S., Tusa, L., Booysen, R., Herrmann, E., Fatihi, A., Möckel, R., Dittrich, T., and Gloaguen, R.: Hyperspectral mineral mapping for underground mining, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-11997, https://doi.org/10.5194/egusphere-egu23-11997, 2023.