Jump to Search Jump to Navigation Jump to Content

AngloGold Ashanti ore sorting

Western Australia

Ore sorting is the generic name for a process that takes a run of mine ore stream and rejects a waste component from further processing. The objective is to reduce the volume of low value material undergoing downstream processing, whilst minimising the misclassification of valuable mineral to the waste stream.

The aim of this project is to conduct a concept study on the use of an automated data analysis technique for ore sorting. Using the particle photographic images and hyperspectral datasets, we will explore the use of various image feature extraction and machine learning methods towards classifying waste from particles of potential value. Cost-sensitive methods will also be explored as a method of fine-tuning the sorting behaviour, from sensitive to conservative.

Ore sorting increases the grade of an ore feed stream by separating very low grade particles (‘waste’) from those containing higher concentrations of the desired mineral (‘ore’), thus economically reducing the amount of material processed in further mineral concentration steps.  This study developed an automated method for discriminating waste and gold-bearing particles using both hyperspectral measurements and RGB images of waste and gold-bearing particles from the Sunrise Dam Gold Mine as input to the discriminating method. Advanced feature extraction methods were employed to capture visual cues such as texture and colour from the RGB images, which were combined with hyperspectral features to give nine types of representative features. Feature selection was applied to groups of the representative features and resulting feature subsets were evaluated using machine learning algorithms to identify a highly informative subset of features. Cost-sensitive training was used to minimise the nominal profit lost due to sorting error based on real cost values from the milling process, with the aim of economically balancing the ore acceptance rate with the waste rejection rate.

Research Outcome:

Horrocks, T., Wedge, D., Holden, E.J., Kovesi, P., Clarke, N., Vann, J., 2015, Classification of gold bearing particles using visual cues and cost sensitive machine learning, Mathematical Geosciences.