Regularized Image Reconstruction from Projections Method

We propose a new image reconstruction from projection method consistent with the maximum a posteriori probability (MAP) approach, i.e. using regularization. We compare reconstruction with three regularization methods -- Tikhonov regularization, smoothed total variation regularization, and early stopping. We compute a number of objective evaluation measures and compare the reconstruction results. We conclude that the best result, both in terms of subjective and objective assessment, is obtained for smoothed total variation regularization.