Abstract:
X-ray Computed Tomography (CT) has been applied in industry for quality and defect control in food products. However, conventional CT systems are neither cost-effective nor flexible, making the deployment of such technology unfeasible for many industrial environments. We propose a simple and cost-effective X-ray imaging setup that comprises a linear translation of the object in a conveyor belt with a fixed X-ray source and detector, with which a small number of X-ray projections can be acquired within a limited angular range. Furthermore, we improved the reconstruction results with the following strategies: (i) an image acquisition involving object rotation during a linear translation in the conveyor-belt; and (ii) an image reconstruction incorporating prior knowledge of the object support (e.g., obtained from optic sensors). Experiments demonstrate the substantial improvement of the reconstruction quality compared to conventional reconstruction methods.
Publications:
L. F. Alves Pereira, Roelandts, T., and Sijbers, J., “Inline 3D X-ray Inspection of Food using Discrete Tomography”, in InsideFood Symposium, Leuven, Belgium, 2013.
L. F. Alves Pereira, Janssens, E., Cavalcanti, G. D. C., Tsang, I. R., Van Dael, M., Verboven, P., Nicolai, B., and Sijbers, J., Inline Discrete Tomography system: application to agricultural product inspection”, Computers and Electronics in Agriculture, vol. 138, pp. 117–126, 2017.
E. Janssens, De Beenhouwer, J., Van Dael, M., Verboven, P., Nicolai, B., and Sijbers, J., “Neural Network-Based X-Ray Tomography for Fast Inspection of Apples on a Conveyor Belt”, in IEEE International Conference on Image Processing, 2015, pp. 917-921.
E. Janssens, De Beenhouwer, J., Van Dael, M., De Schryver, T., Van Hoorebeke, L., Verboven, P., Nicolai, B., and Sijbers, J., “Neural network Hilbert transform based filtered backprojection for fast inline X-ray inspection”, Measurement Science and Technology, vol. 29, no. 3, 2018.