Computed Tomography for automation in food processing

The processing of food products is complex because they are three-dimensional (3D) and microstructured materials. The composition and microstructure of food determine important quality attributes related to taste and texture. Both can vary considerably between and within individual food products. Spatially resolved information of composition and structure of the food will therefore aid and improve processing. To this end, surface (shape) and internal features (structure) are relevant. For each of these, dedicated sensor techniques (such as 3D vision and X-rays) have recently been developed, be it in controlled lab conditions. However, integrated solutions that are compatible with processing line conditions are lacking. In this Ph.D. project, research will be conduct to achieve the implementation of an inline spatially resolved imaging concept to assess internal food quality through differences in internal and external structure and composition. There will be worked with the top-notch X-ray CT imaging facility of KU Leuven and deploy advanced image processing tools, including deep learning approaches, for 3D food structure modeling and dedicated inspection applications. The integrated solution that will be designed (a combination of sensors and advanced processing algorithms) will perform operations that are traditionally done by human operators combining their eyes and brain. Integrated into future automated processing lines the solution will lead to increased resource efficiency and profitability in food production. There will be a collaboration with the technology industry and end-users to test and validate the applications.