Towards a breakthrough in the detection of contaminants in food products

Outlines research work [by the Machine Vision Group at Royal Holloway]into an x‐ray inspection system able to detect contaminants in packaged food products. Describes the analysis of a four‐stage machine vision structure that includes image formation, pre‐processing,feature extraction and decision. The system was tested on various food products and contaminants with the resultant conclusions that the contamination detection system was adaptive and worked with many types of products and contaminants. Refinements in the final decision making stage are being implemented by the use of expert [neural] networks.