A Lagrange method based L-curve for image restoration

3rd International Workshop on New Computational Methods for Inverse Problems

物理学;计算机科学

Landi, G.^1

Department of Mathematics, University of Bologna, Italy^1

关键词: Data noise;    Image restoration problem;    L-Curve;    L-shaped;    Lagrange methods;    Numerical results;    Original images;    Regularization strategies;   

Others  :  https://iopscience.iop.org/article/10.1088/1742-6596/464/1/012011/pdf
DOI  :  10.1088/1742-6596/464/1/012011

学科分类：计算机科学（综合）

来源: IOP

【 摘 要 】

The solution of image restoration problems usually requires the use of regularization strategies. The L-curve criterium is a popular heuristic tool for choosing good regularized solutions, when the data noise norm is not a priori known. In this work, we propose replacing the original image restoration problem with a noise-independent equality constrained one and solving it by an iterative Lagrange method. The sequence of the computed iterates defines a discrete L-shaped curve. By numerical results, we show that good regularized solutions correspond with the corner of this curve.