Utilisation of Enhanced Thresholding for Non-Opaque Mineral Segmentation in Optical Image Analysis

To understand and optimise downstream processing of ores, reliable information about mineral abundance, association, liberation textural characteristics is needed. Such can be obtained by using Optical Image Analysis (OIA) in reflected light, which achieve good discrimination for the majority minerals. However, automated segmentation non-opaque minerals, such as quartz, have reflectivity close to that epoxy they are embedded in, has always been problematic. Application standard thresholding techniques purpose typically results significant misidentifications. This paper presents a sophisticated mechanism, based on enhanced minerals developed Commonwealth Scientific Industrial Research Organisation’s (CSIRO) Mineral5/Recognition5 OIA software, significantly improves many applications. The method utilises an image view adjusted scale more precise initial thresholding, comprehensive clean-up procedures further improvement. For complex cases, also employs specific particle border with subsequent selective erosion-based “reduction borders”, while “particle restoration” prevents detachment grains from larger particles. combined “relief-based minerals” improved overall