Hyperspectral imaging for VIS-SWIR classification of post-consumer plastic packaging products by polymer and color

Improving the quality of recovery and recycling of post-consumer plastic packaging is certainly one of the objectives of the circular economy. Currently, these secondary raw materials still represent only a small part of the materials used in the EU compared to primary ones. Therefore, it is necessary to improve the selection, the separation and recovery techniques with the aim to increase the quantity and quality of these materials put in the market. For these reasons, hyperspectral imaging (HSI) techniques represent a great solution for the characterization, the classification and quality check of different secondary raw materials in several industrial sectors. The present study proposes an efficient characterization of the most used polyolefins, polyethylene (PE) and polypropylene (PP), derived from post-consumer plastic packaging, based on the type of polymer and color, through HSI analysis and the implementation of classification models. Two different HSI acquisition tools were used, working in the short-wave infrared range (1000-2500 nm), to determinate the polymer, and in the visible range (400-700 nm) for the identification by color. In addition, the data processing and the chemometric techniques, used for the development of the classification strategies, have been performed with the PLS_toolbox (Eigenvector Research, Inc.) in Matlab (The MathWorks, Inc.) environment. The obtained results proved the correct identification of the polymer and the color of the investigated plastic flakes, confirming the sustainability and great potential of the HSI analysis techniques, which can be implemented to improve the quality of the plastic materials produced in the recycling plants of polyolefins.