Optical Sorter with Hyperspectral Imaging Techniques for Gluten-Free Beans (Phaseolus Vulgaris)

Abstract

Consumer interest in gluten-free products is growing, as celiac disease (CD) is a chronic autoimmune condition with an increasing global prevalence. CD triggers an immune response upon ingestion of gluten, a protein complex found in cereals such as barley, wheat, and rye.To be classified and certified as gluten-free, a product must be completely free of cereal contaminants. this implies that even trace amounts of gluten-containing grains must be identified and removed to ensure safety for celiac consumers. This research study experimented with the use of optical systems as a hyperspectral image analysis technique to determine and identify some spectral parameters of the analyzed products (beans) with consequent classification of the contaminants (stones, cereals and beans defected). VIS (Visible) and NIR (Near Infrared) cameras present in opposite positions in the sorter allow to acquire 360° hyperspectral images. The integration of a statistical classifier ensures the accurate detection of 100% of contaminants, while the loss of uncontaminated product during sorting is statistically negligible.