HyperSpectral Imaging for the in-line characterization of fine mixtures in WEEE mechanical recycling systems

HyperSpectral imaging (HSI) is a promising technique that could support on-line characterization of multi-material mixtures in recycling. However, this technology currently shows limitations when analyzing metallic mixtures at relatively low particle sizes. In this paper, the set-up of a fast and flexible vision system for the characterization of waste from electric and electronic equipment (WEEE) at fine fractions is presented. The paper deals with technical and chemometric challenges related to: i) the integration of this approach in an on-line de-manufacturing plant; ii) the development of an innovative procedures, based on the analysis of NIR spectral features, for the on-line mixture characterization to perform both sorting of metal and non-metal fractions and the quality control of End Of Life (EOL) products and the different flow streams inside the recycling plant; iii) the application of this technique to very fine products (