A new augmented Lagrangian primal dual algorithm for elastica regularization

被引:7
|
作者
Zhang, Jianping [1 ,2 ]
Chen, Ke [1 ,3 ]
机构
[1] Univ Liverpool, Dept Math Sci, Liverpool L6972L, Merseyside, England
[2] Xiangtan Univ, Sch Math & Computat Sci, Xiangtan, Peoples R China
[3] Univ Liverpool, Ctr Math Imaging Tech, Liverpool L6972L, Merseyside, England
基金
中国国家自然科学基金; 英国工程与自然科学研究理事会;
关键词
Euler's elastica; Legendre-Fenchel transformation; primal-dual technique; augmented Lagrangian method; image denoising; image segmentation;
D O I
10.1177/1748301816668044
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
Regularization is a key element of variational models in image processing. To overcome the weakness of models based on total variation, various high order (typically second order) regularization models have been proposed and studied recently. Among these, Euler's elastica energy based regularizer is perhaps the most interesting in terms of both mathematical and physical justifications. More importantly its success has been proven in applications; however it has been a major challenge to develop fast and effective algorithms. In this paper we propose a new idea for deriving a primal dual algorithm, based on Legendre-Fenchel transformations, for representing the elastica regularizer. Combined with an augmented Lagrangian for-mulation, we are able to derive an equivalent unconstrained optimization that has fewer variables to work with than previous works based on splitting methods. We shall present our algorithms for both the image restoration problem and the image segmentation model. The idea applies to other models where the elastica regularizer is required. Numerical experiments show that the proposed method can produce highly competitive results with better efficiency.
引用
收藏
页码:325 / 338
页数:14
相关论文
共 50 条
  • [1] A primal-dual augmented Lagrangian
    Gill, Philip E.
    Robinson, Daniel P.
    [J]. COMPUTATIONAL OPTIMIZATION AND APPLICATIONS, 2012, 51 (01) : 1 - 25
  • [2] A primal-dual augmented Lagrangian
    Philip E. Gill
    Daniel P. Robinson
    [J]. Computational Optimization and Applications, 2012, 51 : 1 - 25
  • [3] New dual method for elastica regularization
    Song, Jintao
    Pan, Huizhu
    Ding, Jieyu
    Wei, Weibo
    Pan, Zhenkuan
    [J]. PLOS ONE, 2022, 17 (03):
  • [4] A New Penalty Dual-primal Augmented Lagrangian Method and its Extensions
    Ou, Xiaoqing
    Yu, Guolin
    Liu, Jie
    Chen, Jiawei
    Liu, Zhaohan
    [J]. TAIWANESE JOURNAL OF MATHEMATICS, 2024,
  • [5] A primal–dual augmented Lagrangian penalty-interior-point filter line search algorithm
    Renke Kuhlmann
    Christof Büskens
    [J]. Mathematical Methods of Operations Research, 2018, 87 : 451 - 483
  • [6] A globally and quadratically convergent primal-dual augmented Lagrangian algorithm for equality constrained optimization
    Armand, Paul
    Omheni, Riadh
    [J]. OPTIMIZATION METHODS & SOFTWARE, 2017, 32 (01): : 1 - 21
  • [7] New Convergence Properties of the Primal Augmented Lagrangian Method
    Zhou, Jinchuan
    Zhu, Xunzhi
    Pan, Lili
    Zhao, Wenling
    [J]. ABSTRACT AND APPLIED ANALYSIS, 2011,
  • [8] A primal dual modified subgradient algorithm with sharp Lagrangian
    Burachik, Regina S.
    Iusem, Alfredo N.
    Melo, Jefferson G.
    [J]. JOURNAL OF GLOBAL OPTIMIZATION, 2010, 46 (03) : 347 - 361
  • [9] A primal dual modified subgradient algorithm with sharp Lagrangian
    Regina S. Burachik
    Alfredo N. Iusem
    Jefferson G. Melo
    [J]. Journal of Global Optimization, 2010, 46 : 347 - 361
  • [10] A primal-dual augmented Lagrangian penalty-interior-point filter line search algorithm
    Kuhlmann, Renke
    Bueskens, Christof
    [J]. MATHEMATICAL METHODS OF OPERATIONS RESEARCH, 2018, 87 (03) : 451 - 483