Krylov subspace split Bregman methods

被引：1
作者：
Alotaibi, Majed [1 ]
Buccini, Alessandro [2 ]
Reichel, Lothar [1 ]
机构：
[1] Kent State Univ, Dept Math Sci, Kent, OH 44242 USA
[2] Univ Cagliari, Dipartimento Matemat & Informat, I-09124 Cagliari, Italy
来源：
APPLIED NUMERICAL MATHEMATICS | 2023年 / 184卷
关键词：
Split Bregman method; Krylov method; Golub-Kahan bidiagonalization; Fixed point algorithm; Cross validation; REGULARIZATION; PARAMETER; CHOICE;
D O I：
10.1016/j.apnum.2022.10.009
中图分类号：
O29 [应用数学];
学科分类号：
070104 ;
摘要：
Split Bregman methods are popular iterative methods for the solution of large-scale minimization problems that arise in image restoration and basis pursuit. This paper investigates the possibility of projecting large-scale problems into a Krylov subspace of fairly small dimension and solving the minimization problem in the latter subspace by a split Bregman algorithm. We are concerned with the restoration of images that have been contaminated by blur and Gaussian or impulse noise. Computed examples illustrate that the projected split Bregman methods described are fast and give computed solutions of high quality. (c) 2022 IMACS. Published by Elsevier B.V. All rights reserved.
引用
收藏
页码：371 / 390
页数：20
相关论文
共 22 条
  • [1] Bakushinskii A.B., 1984, USSR COMP MATH MATH+, V24, P181
  • [2] Bregman LM., 1967, USSR COMP MATH MATH, V7, P200, DOI [DOI 10.1016/0041-5553(67)90040-7, 10.1016/0041-5553(67)90040-7]
  • [3] Generalized cross validation for lp-lq minimization
    Buccini, Alessandro
    Reichel, Lothar
    [J]. NUMERICAL ALGORITHMS, 2021, 88 (04) : 1595 - 1616
  • [4] Linearized Krylov subspace Bregman iteration with nonnegativity constraint
    Buccini, Alessandro
    Pasha, Mirjeta
    Reichel, Lothar
    [J]. NUMERICAL ALGORITHMS, 2021, 87 (03) : 1177 - 1200
  • [5] Large-scale regression with non-convex loss and penalty
    Buccini, Alessandro
    Cabrera, Omar De la Cruz
    Donatelli, Marco
    Martinelli, Andrea
    Reichel, Lothar
    [J]. APPLIED NUMERICAL MATHEMATICS, 2020, 157 : 590 - 601
  • [6] An lP-lq minimization method with cross-validation for the restoration of impulse noise contaminated images
    Buccini, Alessandro
    Reichel, Lothar
    [J]. JOURNAL OF COMPUTATIONAL AND APPLIED MATHEMATICS, 2020, 375
  • [7] SPLIT BREGMAN METHODS AND FRAME BASED IMAGE RESTORATION
    Cai, Jian-Feng
    Osher, Stanley
    Shen, Zuowei
    [J]. MULTISCALE MODELING & SIMULATION, 2009, 8 (02) : 337 - 369
  • [8] Fast Two-Phase Image Deblurring Under Impulse Noise
    Cai, Jian-Feng
    Chan, Raymond H.
    Nikolova, Mila
    [J]. JOURNAL OF MATHEMATICAL IMAGING AND VISION, 2010, 36 (01) : 46 - 53
  • [9] A SUBSPACE-ACCELERATED SPLIT BREGMAN METHOD FOR SPARSE DATA RECOVERY WITH JOINT l1-TYPE REGULARIZERS
    de Simone, Valentina
    di Serafino, Daniela
    Viola, Marco
    [J]. ELECTRONIC TRANSACTIONS ON NUMERICAL ANALYSIS, 2020, 53 : 406 - 425
  • [10] A new directional weighted median filter for removal of random-valued impulse noise
    Dong, Yiqiu
    Xu, Shufang
    [J]. IEEE SIGNAL PROCESSING LETTERS, 2007, 14 (03) : 193 - 196
123