A Smoothed l0-Norm and l1-Norm Regularization Algorithm for Computed Tomography. (2nd June 2019)

• Record Type: Journal Article
• Authors: Zhu, Jiehua; Li, Xiezhang
• Other Names: Hu Ying Academic Editor.
• Is Part Of: Journal of applied mathematics. Volume 2019(2019)
• Publication Date: 2019-06-02
• Subjects: Mathematics -- Periodicals; 519.05
• DOI: 10.1155/2019/8398035
• Languages: English
• ISSNs: 1110-757X

Abstract

Abstract : The nonmonotone alternating direction algorithm (NADA) was recently proposed for effectively solving a class of equality-constrained nonsmooth optimization problems and applied to the total variation minimization in image reconstruction, but the reconstructed images suffer from the artifacts. Though by thel 0 -norm regularization the edge can be effectively retained, the problem is NP hard. The smoothedl 0 -norm approximates thel 0 -norm as a limit of smooth convex functions and provides a smooth measure of sparsity in applications. The smoothedl 0 -norm regularization has been an attractive research topic in sparse image and signal recovery. In this paper, we present a combined smoothedl 0 -norm andl 1 -norm regularization algorithm using the NADA for image reconstruction in computed tomography. We resolve the computation challenge resulting from the smoothedl 0 -norm minimization. The numerical experiments demonstrate that the proposed algorithm improves the quality of the reconstructed images with the same cost of CPU time and reduces the computation time significantly while maintaining the same image quality compared with thel 1 -norm regularization in absence of the smoothedl 0 -norm.