Advancing Plastic Recycling with Computer Vision
Efficient recycling of plastic waste requires accurate identification of material composition, an ongoing challenge, particularly for complex or dark-coloured plastics widely used in industries such as automotive manufacturing. Within the CIRCULOOS project, a computer vision–based system for composition detection has been developed to address this issue and support more effective circular manufacturing processes.
The system combines hyperspectral imaging, spectroscopy techniques, and AI algorithms to identify plastics based on their unique spectral signatures. By analysing how different materials reflect light across multiple wavelengths, the technology can classify plastic types and support automated sorting processes, enabling higher-quality recycled materials to be reintroduced into production cycles.
Designed for integration into industrial recycling environments, this solution improves waste classification, reduces material contamination, and strengthens circular supply chains—helping manufacturers transform plastic waste into valuable resources.