Analysis of micro coastal litter in the visible and infrared bands through the use of a spectrometer and a hyperspectral camera

Abstract

Continuous and extensive plastic use, coupled with inadequate waste management, has resulted in the pervasive existence of plastic pollution in the marine environment. Of particular concern is the widespread distribution of microplastics, posing a significant threat to marine life and public health. The absence of a standardised approach for microplastic analysis has been evident, prompting significant efforts within the scientific community. Traditional methods have proven time-consuming and impractical for standardisation. Establishing a standardised method is essential for comprehensive and precise result comparisons, yielding dependable data and conclusions, while being fast, non-invasive, and superior to existing alternatives. This study aims to assess the efficiency of a hyperspectral camera in accurately identifying microplastics, specifically their polymer type. A hand-held near-infrared (NIR) spectrometer, known for its accuracy despite slower results, served as the comparative method. Forty microplastic samples were collected from a beach in Malta and investigated using both a NIR spectrometer and a hyperspectral camera. The NIR spectrometer successfully identified four polymer types: polyethylene (PE), polypropylene (PP), polyvinyl chloride (PVC), and polyolefins (PO), producing unique spectral signatures for each sample. These spectral signatures were then compared to those generated by the hyperspectral camera, which exhibited minimal variability and distinct, prominent spectral patterns for each of these polymer type. This study demonstrates the capability of hyperspectral imaging (HSI) to efficiently and accurately establish a baseline for rapid microplastic identification, thereby highlighting its potential as a standardised method and the need for developing a spectral library.