Inline 3D X-ray Inspection of Food using Discrete Tomography

While X-ray radiography is commonly used for inline inspection of food, it is insufficient if 3D information is needed from the scanned object. Computed Tomography (CT) on the other hand, does provide 3D information, but is known to be expensive and relatively slow. This work proposes a low-cost X-ray scanning system based on a static X-ray source with a wide cone, along with a large detector with which only few projections are acquired while food passes by on a conveyor belt. For such scanning geometry, conventional tomographic reconstruction methods would lead to images that are severely polluted with smearing artefacts. In this paper, we propose the use of the Discrete Algebraic Reconstruction Technique (DART) to reduce these artefacts. Simulation results show that DART can generate reconstructions of much higher quality than conventional reconstruction methods in conveyor belt X-ray scanning which is an important step towards quantitative analysis of 3D objects for inline Xray inspection.