Iterative reconstruction of coded source neutron radiographs

Use of a coded source facilitates high-resolution neutron imaging but requires that the radiographic data be deconvolved. In this paper, we compare direct deconvolution with two different iterative algorithms, namely, one based on direct deconvolution embedded in an MLE-like framework and one based on a geometric model of the neutron beam and a least squares formulation of the inverse imaging problem. Simulation results show that the model-based reconstruction method is the suitable solution for high-resolution coded source systems and where the illumination of the neutron source is non-uniform.