Optical Mineral Classification by Acquisition and Processing of Hyperspectral Fluorescence Images

摘要

In the search of fast, reliable and cost-efficient methods for industrial mineral sorting, new methods are being investigated. We present a yet untried method, namely the analysis of mineral fluorescence by means of hyperspectral imaging. For this method, minerals are excited by ultraviolet radiation and their space-resolved visible fluorescence light is recorded spectrally. We present the hyperspectral images as well as improved classification results obtained by the usage of denoised images.