Hyperspectral imaging raising the bar for plastics sorting

Canadian tech firm Metaspectral has won a contract of almost US$ 110 000 (EUR 102 000) to tackle plastic film and flexible packaging.

Environment Climate Change Canada has commissioned Metaspectral to demonstrate its artificial intelligence sorting technology. The contract is part of the Canadian Plastics Innovation Challenge.

Metaspectral, which was launched in Vancouver in 2018, relies on specialised cameras to capture hyperspectral images of plastic scrap. This allows the tech provider to detect the materials’ chemical signatures – in real time, pixel by pixel. Deep learning algorithms process the images, identifying the composition of the materials at molecular level.

All polymers recognised

Metaspectral’s vision system for the recycling industry is designed to be installed on top of any conveyor belt. It detects all types of polymers, no matter whether the material is wet, dirty, perforated, etc. Positive results have been proven for transparent, black, or mixed polymers.  

Operators can customise Metaspectral’s vision system according to the task at hand. They have the capability to gain visibility in the NIR (900 – 1700 nm), SWIR (1000 – 2500 nm) or MWIR (2700 – 5300 nm) ranges of the electromagnetic spectrum.

Metaspectral is even working together with the Canadian Space Agency to help better understand the Earth’s surface.