Technology

Advanced material detection via AI-based hyperspectral imaging

We believe that a true solution to textile waste is going to require innovative applications of technology and cutting-edge processes.

That’s why we’ve developed a highly accurate system from the ground-up that powers detection and sorting across a broad range of fibers.

Partner with Refiberd

How it works

1
The textile is placed underneath the hyperspectral camera and the line light source.
2

Camera measures lines of hyperspectral data from the moving textile at a specified framerate.

Computer stitches together lines of hyperspectral data to create a hyperspectral cube such that each pixel of the image is a spectra.

3
The hyperspectral cube is processed by a machine learning model. The model outputs a prediction of the material composition of the textile.
Our technology combines a hyperspectral imaging system with artificial intelligence to accurately detect fiber composition and contaminant presence in textile waste.

Advanced hyperspectral imaging system

Hyperspectral imaging detects the way light interacts with material - different materials absorb and reflect light differently based on their chemical composition - allowing for identification across materials.

Our state of the art hyperspectral imaging system is tuned for the most sensitive detection of textile fiber types, able to detect even trace amounts of material and contamination.

Combined with artificial intelligence

Our artificial intelligence, powered by a proprietary dataset of thousands of custom textile samples, then processes the hyperspectral imaging data to identify all the different fibers present in waste, including blended and layered fibers.

The result is highly-accurate imaging and smarter detection across a broad range of textile types.

Our team of technologists