Research articles (Dept of Geology)http://hdl.handle.net/10311/772024-03-28T20:02:30Z2024-03-28T20:02:30ZCoupling NCA dimensionality reduction with machine learning in multispectral rock classification problemsSinaice, Brian BinoOwada, NarihiroSaadat, MahdiToriya, HisatoshiInagaki, FumiakiBagai, ZibisaniKawamura, Youheihttp://hdl.handle.net/10311/24852023-01-17T00:02:57Z2021-08-05T00:00:00ZCoupling NCA dimensionality reduction with machine learning in multispectral rock classification problems
Sinaice, Brian Bino; Owada, Narihiro; Saadat, Mahdi; Toriya, Hisatoshi; Inagaki, Fumiaki; Bagai, Zibisani; Kawamura, Youhei
Though multitudes of industries depend on the mining industry for resources, this industry has taken hits in terms of declining mineral ore grades and its current use of traditional, time consuming and computationally costly rock and mineral identification methods. Therefore, this paper proposes integrating Hyperspectral Imaging, Neighbourhood Component Analysis (NCA) and Machine Learning (ML) as a combined system that can identify rocks and minerals. Modestly put, hyperspectral imaging gathers electromagnetic signatures of the rocks in hundreds of spectral bands. However, this data suffers from what is termed the ‘dimensionality curse’, which led to our employment of NCA as a dimensionality reduction technique. NCA, in turn, highlights the most discriminant feature bands, number of which being dependent on the intended application(s) of this system. Our envisioned application is rock and mineral classification via unmanned aerial vehicle (UAV) drone technology. In this study, we performed a 204-hyperspectral to 5-band multispectral reduction, because current production drones are limited to five multispectral bands sensors. Based on these bands, we applied ML to identify and classify rocks, thereby proving our hypothesis, reducing computational costs, attaining an ML classification accuracy of 71%, and demonstrating the potential mining industry optimisations attainable through this integrated system.
2021-08-05T00:00:00ZContaminant binding and bioaccessibility in the dust from the Ni-Cu mining/smelting District of Selebi-Phikwe (Botswana)Ettler, VojtěchHladíková, KarolínaMihaljevič, MartinDrahota, PetrCulka, AdamJedlicka, RadimKříbek, BohdanSracek, OndraBagai, Zibisanihttp://hdl.handle.net/10311/24842023-01-17T00:01:49Z2022-08-22T00:00:00ZContaminant binding and bioaccessibility in the dust from the Ni-Cu mining/smelting District of Selebi-Phikwe (Botswana)
Ettler, Vojtěch; Hladíková, Karolína; Mihaljevič, Martin; Drahota, Petr; Culka, Adam; Jedlicka, Radim; Kříbek, Bohdan; Sracek, Ondra; Bagai, Zibisani
We studied the dust fractions of the smelting slag, mine tailings, and soil from the former Ni-Cu mining and processing district in Selebi-Phikwe (eastern Botswana). Multi-method chemical and mineralogical
investigations were combined with oral bioaccessibility testing of the fine dust fractions (<48 and <10 μm)
in a simulated gastric fluid to assess the potential risk of the intake of metal(loid)s contaminants. The total
concentrations of the major contaminants varied significantly (Cu: 301–9,600 mg/kg, Ni: 850–7,000 mg/kg,
Co: 48–791 mg/kg) but were generally higher in the finer dust fractions. The highest bioaccessible
concentrations of Co, Cu, and Ni were found in the slag and mine tailing dusts, where these metals were mostly
bound in sulfides (pentlandite, pyrrhotite, chalcopyrite). On the contrary, the soil dusts exhibited substantially
lower bioaccessible fractions of these metals due to their binding in less soluble spinel-group oxides. The results indicate that slag dusts are assumed to be risk materials, especially when children are considered as a target group. Still, this exposure scenario seems unrealistic due to (a) the fencing of the former mine area and its inaccessibility to the local community and (b) the low proportion of the fine particles in the granulated slag
dump and improbability of their transport by wind. The human health risk related to the incidental ingestion of
the soil dust, the most accessible to the local population, seems to be quite limited in the Selebi-Phikwe area,
even when a higher dust ingestion rate (280 mg/d) is considered.
2022-08-22T00:00:00ZSpectral angle mapping and AI methods applied in automatic identification of placer deposit magnetite using multispectral camera mounted on UAVSnaice, Brian BinoOwada, NarihiroIkeda, HajimeToriya, HisatoshiBagai, ZibisaniShemang, ElishaAdachi, TsuyoshiKawamura, Youheihttp://hdl.handle.net/10311/24832023-01-17T00:01:24Z2022-02-20T00:00:00ZSpectral angle mapping and AI methods applied in automatic identification of placer deposit magnetite using multispectral camera mounted on UAV
Snaice, Brian Bino; Owada, Narihiro; Ikeda, Hajime; Toriya, Hisatoshi; Bagai, Zibisani; Shemang, Elisha; Adachi, Tsuyoshi; Kawamura, Youhei
The use of drones in mining environments is one way in which data pertaining to the state of a site in various industries can be remotely collected. This paper proposes a combined system that employs a 6-bands multispectral image capturing camera mounted on an Unmanned Aerial Vehicle (UAV) drone, Spectral Angle Mapping (SAM), as well as Artificial Intelligence (AI). Depth possessing multispectral data were captured at different flight elevations. This was in an attempt to find the best elevation where remote identification of magnetite iron sands via the UAV drone specialized in collecting spectral information at a minimum accuracy of +/- 16 nm was possible. Data were analyzed via SAM to deduce the cosine similarity thresholds at each elevation. Using these thresholds, AI algorithms specialized in classifying imagery data were trained and tested
to find the best performing model at classifying magnetite iron sand. Considering the post flight logs, the spatial area coverage of 338 m2, a global classification accuracy of 99.7%, as well the per-class precision of 99.4%, the 20 m flight elevation outputs presented the best performance ratios overall. Thus, the positive outputs of this study suggest viability in a variety of mining and mineral engineering practices.
This article is an expanded version this conference paper: Sinaice, B.B.; Takanohashi, Y.; Owada, N.; Utsuki, S.;
Hyongdoo, J.; Bagai, Z.; Shemang, E.; Kawamura, Y. Automatic magnetite identification at Placer deposit
using multi-spectral camera mounted on UAV and machine learning. In Proceedings of the 5th International
Future Mining Conference 2021—AusIMM 2021, Online, 6–10 December 2021; pp. 33–42;
ISBN 978-1-922395-02-3.
The symbols may not appear as in the original article
2022-02-20T00:00:00ZUse of the geochemical and biological sedimentary record in establishing palaeo-environments and climate changein the Lake Ngami basin, NW BotswanaHuntsman-Mapila, P.Ringrose, S.Mackay, A.W.Downey, W.S.Modisi, M.Coetzee, S.H.Tiercelin, J.-J.Kampunzu, A.B.Vanderpost, C.http://hdl.handle.net/10311/8912016-08-13T00:10:38Z2006-01-01T00:00:00ZUse of the geochemical and biological sedimentary record in establishing palaeo-environments and climate changein the Lake Ngami basin, NW Botswana
Huntsman-Mapila, P.; Ringrose, S.; Mackay, A.W.; Downey, W.S.; Modisi, M.; Coetzee, S.H.; Tiercelin, J.-J.; Kampunzu, A.B.; Vanderpost, C.
Sediment samples from a continuous 4.6m profile in the dry bed of Lake Ngami in NW Botswana were analysed for geochemistry and
dated using both 14C and TL methods. Certain units in the profile were found to be diatom rich and these, with the geochemical results,
were used as indicators of high and low lake levels within the basin. The Lake Ngami sediments contain a high proportion of SiO2 (51–92.5 wt%, avg. 72.4 wt%) and variable levels of Al2O3 (2.04–17.2 wt%, avg. 8.88 wt%). Based on elevated Al2O3 and organic matter
(LOIorgC) results, lacustrine conditions occurred at ca. 42 ka until 40 ka and diatom results suggest that relatively deep but brackish
conditions prevailed. At 40 ka, the lacustrine sedimentary record was terminated abruptly, possibly by tectonic activity. At ca. 19 ka, shallow, aerobic, turbulent conditions were prevalent, but lake levels were at this time increasing to deeper water conditions up until ca. 17 ka. This period coincides with the Late Glacial Maximum, a period of increased aridity in the central southern Africa region.
Generally, increasing Sr/Ca ratios and decreasing LOIorgC and Al2O3, from ca. 16 to 5 ka, suggest decreasing inflow into the basin and
declining lake levels. Based on the enrichment of LREE results, slightly alkaline conditions prevailed at ca. 12 ka. Diatom results also support shallow alkaline conditions around this time. These lake conditions were maintained primarily by local rainfall input as the region experienced a warmer, wetter phase between 16 and 11 ka. Lake levels rose rapidly by 4 ka, probably in response to enhanced rainfall in the Angolan catchment. These results indicate that lake levels in the Lake Ngami basin are responding to rainfall changes in
the Angolan catchment area and local rainfall. The results confirm that the present-day anti-phase rainfall relationship between southern
Africa and regions of equatorial Africa was extant during the late Quaternary over the Angolan highlands and NW Botswana.
2006-01-01T00:00:00Z