Methods of filtering densities reconstructed from line projections

Abstract

We present details important for practical applications of a new algorithm for filtering densities reconstructed from their line projections. The method is applied to densities reconstructed via the Cormack method from both model projections and 2D angular correlation of positron annihilation radiation spectra in CeIn3 and ErGa3. It is shown that this algorithm filters the experimental noise as well as enhances some details of the reconstructed densities, which then reveals their fine features. We also compare this way of filtering with the effects of imposing symmetry conditions on densities reconstructed for cubic systems.