In mining operations, an ore is separated into its constituents through mineral processing methods, such as flotation. Identifying the type of minerals contained in the ore in advance aids greatly in performing faster and more efficient mineral processing. The human eye can recognize visual information in three wavelength regions: red, green, and blue. With hyperspectral imaging, high resolution spectral data that contains information from the visible light wavelength region to the near infrared region can be obtained. Using deep learning, the features of the hyperspectral data can be extracted and learned, and the spectral pattern that is unique to each mineral can be identified and analyzed. In this paper, we propose an automatic mineral identification system that can identify mineral types before the mineral processing stage by combining hyperspectral imaging and deep learning. By using this technique, it is possible to quickly identify the types of minerals contained in rocks using a non-destructive method. As a result of experimentation, the identification accuracy of the minerals that underwent deep learning on the red, green, and blue (RGB) image of the mineral was approximately 30%, while the result of the hyperspectral data analysis using deep learning identified the mineral species with a high accuracy of over 90%.
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.
Fragmentation size distribution estimation is a critical process in mining operations that employ blasting. In this study, we aim to create a low-cost, efficient system for producing a scaled 3D model without the use of ground truth data, such as GCPs (Ground Control Points), for the purpose of improving fragmentation size distribution measurement using GNSS (Global Navigation Satellite System)-aided photogrammetry. However, the inherent error of GNSS data inhibits a straight-forward application in Structure-from-Motion (SfM). To overcome this, the study proposes that, by increasing the number of photos used in the SfM process, the scale error brought about by the GNSS error will proportionally decrease. Experiments indicated that constraining camera positions to locations, relative or otherwise, improved the accuracy of the generated 3D model. In further experiments, the results showed that the scale error decreased when more images from the same dataset were used. The proposed method is practical and easy to transport as it only requires a smartphone and, optionally, a separate camera. In conclusion, with some modifications to the workflow, technique, and equipment, a muckpile can be accurately recreated in scale in the digital world with the use of positional data.
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.
In this paper, the local correspondence between synthetic aperture radar (SAR) images and optical images is proposed using an image feature-based keypoint-matching algorithm. To achieve accurate matching, common image features were obtained at the corresponding locations. Since the appearance of SAR and optical images is different, it was difficult to find similar features to account for geometric corrections. In this work, an image translator, which was built with a DNN (deep neural network) and trained by conditional generative adversarial networks (cGANs) with edge enhancement, was employed to find the corresponding locations between SAR and optical images. When using conventional cGANs, many blurs appear in the translated images and they degrade keypoint-matching accuracy. Therefore, a novel method applying an edge enhancement filter in the cGANs structure was proposed to find the corresponding points between SAR and optical images to accurately register images from different sensors. The results suggested that the proposed method could accurately estimate the corresponding points between SAR and optical images.
The adoption of hyperspectral imaging has had positive feedback in multiple industries, especially those heavily reliant on the visual analysis of subjects. Reasons for such are primarily due to the high accuracies achievable from processing high dimensional data. Nevertheless, hyperspectral data is said to possess a ' dimensionality curse'. This phenomenon, deems it computationally demanding and diffi cult to employ in rapid fi eld investigations such as the use of drone-mounted spectral cameras to distinguish rocks. To counter this, this study proposes the employment of a method of reducing the number of dimensions used to highlight the most characteristic feature bands referred to as Neighbourhood Component Analysis(NCA). NCA aided in disregarding redundant bands from 204 dimensionalities, to a still highly capable 5 bands dimensionality, which coincides with the current production of 5-band detection drones. To process this data, several machine learning(ML) algorithms were run in order to perform spectral classifi cation of rocks based on the 5 NCA defi ned bands. This study's novel fi ndings show that one is able to acquire with NCA and ML, 5 bands, with a post-optimization average global accuracy of 95.4%. Such capabilities are highly suffi cient considering the magnitude of the dimensionality reduction combined with the potential fi eld drone applicability.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.