Applicability of near - infrared hyperspectral imagery (NIR-HI) for sensor based sorting of an epithermal AU-AG ore

Abstract

In the presented study test work was performed with near-infrared hyperspectral imagery (NIR-HI) on 36 ore samples from a South-American epithermal Au-Ag mine. The aim of the test work was to investigate if NIR-HI provides information about the alteration mineralogy of samples that can be used to predict their economic value. Mineral distribution maps were produced from the hyperspectral images by using correlation coefficients between the image pixels and a set of reference spectra. These maps showed that detection of mineralogy with NIR-HI can be used to distinguish; i) ore particles with low Au and Ag grades, ii) ore particles with high carbon contents, iii) ore particles with high sulphur contents.