Monitoring Of Micro-Plastics From Marine Environment Adopting HyperSpectral Imaging

The presence of micro-plastic debris in the marine environment contaminates oceans, producing dangerous effects. Micro-plastics, due to their small size, enter in the marine food chain, causing harmful effects on the whole ecosystem. The possibility to develop a sensor based procedure to monitor marine micro-plastics was explored in this work. More in detail, this study was addressed to detect and to recognize different types of micro-plastics coming from sampling in different sea areas (i.e. offshore, on the coast) adopting an approach based on HyperSpectral Imaging (HSI). This approach is based on the utilization of a device working in the short wave infrared range (SWIR) from 1000 to 2500 nm: a Specim SISUChema XL™, embedding an ImSpector™ N25E (Specim Ltd, Finland), was utilized to collect the hyperspectral data. Acquired hyperspectral images were analyzed adopting the PLS_Toolbox (Version 8.1, Eigenvector Research, Inc.) running into the Matlab® environment (Version 8.6, The Mathworks, Inc.). Preliminary analyses were performed starting from the collection of information acquired inside selected Regions Of Interest (ROI). The resulting collected spectral features of each particle were preliminary analysed by visual inspection. Afterwards, starting from spectral libraries of virgin plastics, assumed as references, the recognition of the different polymers was performed, adopting different classification methods and comparing the results. In order to perform a full characterization of the samples, morphological and morphometrical particle analysis was also carried out. The results achieved from this study can represent an important starting point to develop, implement and set up monitor strategies to characterize marine micro-plastics.