Recycling-oriented characterization of PET waste stream by SWIR hyperspectral imaging and variable selection methods

The proposed study was carried out to develop a fast and efficient strategy for plastic waste sensor-based sorting in recycling plants, based on hyperspectral imaging (HSI), to produce a high-quality recycled polyethylene terephthalate (PET) stream, in agreement with the principles of circular economy. Variable selection techniques were applied in order to identify a limited number of spectral bands useful to recognize the different classes of materials, reducing processing time as requested by sorting online applications. Post-consumer plastic samples were analyzed in order to detect the presence of other plastic materials, considered contaminants, inside a stream of recycled PET flakes. Hyperspectral images of plastic flakes were acquired by HSI working in the short-wave infrared (SWIR) range (1000 - 2500 nm). As a first step, the hypercubes were processed applying chemometric logics to build a classification model, based on the full investigated spectral range, able to identify PET and contaminants. As a secondstep, different variable selection methods were then applied in order to identify the fundamental spectral bands useful to recognize the two classes and to evaluate the best method, showing efficiency valuesclose to those obtained by the full spectrum model. The obtained results suggested that the variables selection approach can represent a powerful approach for PET sorting, decreasing the amount of data tobe processed and thus enabling faster recognition compared to the full spectrum model. Therefore, the developed non-destructive method based on HSI can be used for the implementation of a quick and effective selection logic of PET flakes running in a sensor-based sorting unit.