Automatic identification and classification of compostable and biodegradable plastics using hyperspectral imaging

In the UK waste management systems biodegradable and compostable packaging are not automatically detected separated. As a result, their fate is generally landfill or incineration, neither of which an environmentally good outcome. Thus, effective sorting technologies for plastics needed to help improve composting rates these materials reduce contamination recycling streams. Hyperspectral imaging (HSI) was applied in this study develop classification models identifying classifying with analysis focused on spectral region 950–1,730 nm. The experimental design includes hyperspectral camera, allowing different chemometric techniques be including principal component (PCA) partial least square discriminant (PLS-DA) model plastics. Materials used included (sugarcane-derived palm leaf derived), (PLA, PBAT) conventional (PP, PET, LDPE). Our strategy identify categorize various fragments over size range 50 x mm 5 mm. Results indicated that both PCA PLS-DA achieved scores 100% when material larger than 10 However, misclassification rate increased 20% sugarcane-derived 40% leaf-based at sizes below. addition, mm, LDPE PBAT 20%, sugarcane palm-leaf based 60 80% respectively while PLA, PP, PET still 0%. system capable accurately (compostable spoons, forks, coffee lids) differentiating them from identical looking plastic items high accuracy.