Abstract
Sensor technologies provide relevant information on the key geological attributes in mining. The integration of data from multiple sources is advantageous in making use of the synergy among the outputs for the enhanced characterisation of materials. Sensors produce various types of data. Thus, the fusion of these data requires innovative data-driven strategies. In the present study, the fusion of image and point data is proposed, aiming for the enhanced classification of ore and waste materials in a polymetallic sulphide deposit at 3%, 5% and 7% cut-off grades. The image data were acquired in the visible-near infrared (VNIR) and short-wave infrared (SWIR) regions of the electromagnetic spectrum. The point data cover the mid-wave infrared (MWIR) and long-wave infrared (LWIR) spectral regions. A multi-step methodological approach was developed for the fusion of the image and point data at multiple levels using the supervised and unsupervised classification techniques. Several possible combinations of the data blocks were evaluated to select the optimal combinations in an optimised way. The obtained results indicate that the individual image and point techniques resulted in a successful classification of ore and waste materials. However, the classification performance greatly improved with the fusion of image and point data, where the K-means and support vector classification (SVC) models provided acceptable results. The proposed approach enables a significant reduction in data volume while maintaining the relevant information in the spectra. This is principally beneficial for the integration of data from high-throughput and large data volume sources. Thus, the effectiveness and practicality of the approach can permit the enhanced separation of ore and waste materials in operational mines.
Highlights
The mining industry relies on access to accurate data on the key geological attributes, as reliable information is a central enabler for effective process control and decision-making in commercial mines.Sensor technologies allow obtaining relevant and accurate information on the geological attributes that are crucial in mining
This study proposes a method of combining the hyperspectral images (VNIR and short-wave infrared (SWIR)) with point spectrometer (MWIR and long-wave infrared (LWIR)) data using machine learning techniques and data fusion approaches for the separation of sulphide ore from waste materials with no diagnostic absorption features of the sulphide minerals
Different scenarios were investigated to assess the use of image and point data integration for the discrimination of ore and waste materials in polymetallic sulphide deposits using infrared technologies, namely: (1)
Summary
Sensor technologies allow obtaining relevant and accurate information on the geological attributes that are crucial in mining (e.g., mineralogy, geochemistry and ore–waste ratio). Such information can be generated along the mining value chain (e.g., during material extraction, transport and processing). Accurate information potentially leads to increased productivity, efficient data management, and enables one to ensure the compliance of the mining process with environmental safety standards. The dynamic development of sensors resulted in high-end technologies that are rapid and efficient for the accurate characterisation of various types of materials.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
