FTVd is beyond Fast Total Variation regularized Deconvolution

TL;DR

This paper analyzes the FTVd algorithm for total variation deconvolution, revealing that intermediate iterative results can outperform the final solution in image quality due to their relation to combined regularization models.

Contribution

It demonstrates that intermediate results in FTVd iterations correspond to combined Tikhonov and total variation models, often yielding better images than the final solution.

Findings

Intermediate results can outperform final solutions in image quality.

Intermediate solutions correspond to combined regularization models.

The analysis provides insights into the iterative process of FTVd.

Abstract

In this paper, we revisit the "FTVd" algorithm for Fast Total Variation Regularized Deconvolution, which has been widely used in the past few years. Both its original version implemented in the MATLAB software FTVd 3.0 and its related variant implemented in the latter version FTVd 4.0 are considered \cite{Wang08FTVdsoftware}. We propose that the intermediate results during the iterations are the solutions of a series of combined Tikhonov and total variation regularized image deconvolution models and therefore some of them often have even better image quality than the final solution, which is corresponding to the pure total variation regularized model.