An iterative procedure for detecting foreign products in powder foods in NIR hyperspectral imaging spectroscopy: frauds and contaminations

Food adulteration is an issue of growing concern, specially delicate throughout early stages of food processing, regarding the control of the original ingredients included in the process and crossed contaminations. The use of Hyperspectral Imaging (HSI) to detect nuts contamination in powder food products is explored in this work. This study deals with two specific problems of using HSI for the detection of food adulterants: the low dose at which they need to be detected and a high similarity between the spectral NIR signatures of adulterating substances. A chemometric method, issued from a pharmaceutical context, is adapted and tested to detect and identify the presence of different species of oily seeds and nuts in a specific type of commercial wheat flour. The proposed approach was able to detect the presence of 5 kinds of contaminants in percentages from 1 to 5%, but not to identify the contaminant species.