Global plastic production is increasing at an alarming rate, expected to double by 2050, and currently accounts for 4.5% of worldwide greenhouse gas emissions. This trend is driving governments and researchers to develop faster and more efficient methods for sorting and recycling plastics to reduce environmental impact.
Effective recycling of plastics requires precise separation of mixed plastic waste by polymer type. Traditional sorting methods, such as color and transparency, often fall short in accuracy. Hyperspectral imaging technology offers a more precise solution, making it an essential part of the sorting process by providing detailed spectral information for material identification.
More information about Headwall's VNIR hyperspectral sensor
Key polymers in plastic products include PE (polyethylene), PET (polyethylene terephthalate), PS (polystyrene), PP (polypropylene), PA (polyamide or nylon), and PVC (polyvinyl chloride). These materials are used in everyday items such as drink cups, food containers, bottles, building materials, and textiles. For effective recycling, hyperspectral imaging can sort materials both before and after shredding, ensuring high-quality material separation.
Image by frimufilms on Freepik
Plastic material is washed and then conveyed along a belt under hyperspectral imaging systems, which analyze the reflected light. Chemometric classification models are applied to each pixel or object, allowing for precise identification. The results are then sent to a Programmable Logic Controller (PLC) to perform sorting based on polymer composition.
For detailed color sorting, VNIR sensors (400 - 1000 nm) are highly effective, significantly outperforming conventional RGB cameras. However, the NIR wavelength range (900 - 1700 nm) is ideal for distinguishing the “spectral fingerprints” of various polymers. Machine learning and statistical algorithms allow for real-time material recognition and sorting, enhancing the accuracy of plastic recycling.
A recent study utilized a hyperspectral imaging system by inno-spec GmbH (now part of the Headwall Group), featuring a RedEye 1.7 camera (950 - 1700 nm), a compact scanning table, halogen tungsten lighting, and a host computer with PerClass Mira Software. Known plastic samples were scanned to create an accurate classification model, which was then applied to mixed plastic batches. The system successfully detected and classified plastics in real-time, simulating an industrial setting.
Black plastic presents a unique challenge due to its lack of reflected light in the VNIR and NIR spectra. In these cases, the BlackEye hyperspectral sensor, with a wavelength range of 2900 - 4200 nm (2.9 µm - 4.2 µm), is ideal for precise detection of black plastics, overcoming limitations of traditional sorting methods.
Image by Freepik
Contact our specialist in spectroscopy Kees van der Sar.
More information on Real-Time, Contactless Sorting of Plastics