Thermographic techniques and adapted algorithms for automatic detection of foreign bodies in food

摘要

At the beginning of the paper, the different methods of capturing thermographic images like pulse and online thermography are explained. Afterwards, the capability of these techniques to differentiate between foreign bodies and food is shown in different examples like hazelnuts contaminated with rotten nuts, stones and hard shells or winegums spoilt with wood particles. For an automatic detection and separation of foreign bodies from a food stream, different specially adapted algorithms are needed. Finally, different techniques like heat dissipation, histogram analysis, texture analysis and an object oriented method are compared to show the quality and reliability of the different methods applied to different materials.