Hyperspectral image analysis combines the characteristics of image processing with spectrography. A hyperspectral image contains greater information than a conventional image even in non-visible wavelengths. This can help to distinguish objects that visually seem identical or to analyse different chromatic variations.
Tecnalia has developed algorithms for the classification of materials based on their hyperspectral images. This system can classify different elements based on a learning of simple samples and employing regression algorithms that allow the characterization of the chemical composition of the materials in the images.
These algorithms have been used in the classification of materials in several industries, such as: recycling, slag classification, forest classification based on aerial images, the food industry or chromatography and colour characterization, to name a few.