Near InfraRed-based hyperspectral imaging approach for secondary raw materials processing in solid waste sector

In secondary raw materials and industrial recycling sectors there is the need of solving quality control issues. The development and deployment of an effective, fast and robust sensing architecture able to detect, characterize and sort solid waste products is of primary importance. Near InfraRed (NIR) based HyperSpectral Imaging (HSI) techniques to detect materials to recycle and/or solid waste products to process represents an interesting solution to address quality control issues in these sectors. In this paper, are presented two different case studies on the utilization of NIR-HSI to detect contaminants in household plastic packaging waste and recognize materials occurring in processed monitors and flat screen waste. The proposed approach consists of a cascade detection based on Partial Least Squares – Discriminant Analysis (PLS-DA) classifiers applied on hyprspectral images acquired in NIR range (1000-1700 nm).