Computed Tomography for Plasma Measurement

Image Reconstruction from Sparse Data

Abstract

Computed tomography studies for plasma imaging are reviewed from a standpoint of numerically coping the sparseness in projection measurement. After a brief note on the failure of the standard methods of reconstruction, the methods analytical with series expansions and algebraic with discrete formulations are summarized with an aspect of linear and nonlinear regularizations of least squares solutions.The usefulness of neg-entropies like AIC and GCV for improving the obtained image is emphasized with successful examples in experiment.