Elastic bending total variation model for image inpainting with operator splitting method

被引:0
作者
Nan, Caixia [1 ]
Zhang, Qian [2 ]
机构
[1] Hong Kong Polytech Univ, Dept Appl Math, Kowloon, Hong Kong, Peoples R China
[2] Harbin Inst Technol, Sch Sci, Shenzhen 518055, Peoples R China
关键词
Image inpainting; Elastic bending energy; Total variation regularization; Operator splitting method; REGULARIZATION; APPROXIMATIONS; SEGMENTATION; EFFICIENT; ADHESION; SCHEMES;
D O I
10.1016/j.camwa.2024.09.023
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
The elastic bending energy model is commonly used to describe the shape transformation of biological lipid vesicles, making it a classical phase field model. In this paper, by coupling the elastic bending energy with the total variation (TV) regularization, we develop an elastic bending-TV model for image inpainting. By solving the energy minimization problem of this model, we obtain the results for image processing. We adopt an operator splitting method for the model and the numerical scheme involves the introduction of two vector- and scalar- valued functions to reconstruct this functional. The energy minimization problem is transformed into finding the steady state solution of artificial time-dependent PDE systems. At each fractional step, we can find either a closed- form solution or being solved by an efficient algorithm, which is a very robust and stable algorithm. Experimental results validate the superiority of our model and the effectiveness of the scheme for image inpainting.
引用
收藏
页码:150 / 164
页数:15
相关论文
共 50 条
  • [31] Image Inpainting Method Based Sparse Analysis Model Of Synchronous Dictionary Learning
    Li, Bin
    Sun, Baohua
    Li, Dekun
    Jiang, Tong
    Li, Gang
    Li, Hao
    2021 INTERNATIONAL CONFERENCE ON IMAGE, VIDEO PROCESSING, AND ARTIFICIAL INTELLIGENCE, 2021, 12076
  • [32] An image inpainting method based on a convex variant of the Mumford-Shah model
    Zhi, Zhanjiang
    Sun, Yi
    Journal of Information and Computational Science, 2015, 12 (11): : 4349 - 4356
  • [33] A novel image inpainting method based on a modified Lengyel-Epstein model
    Wang, Jian
    Luo, Mengyu
    Chen, Xinlei
    Xu, Heming
    Kim, Junseok
    COMPUTER VISION AND IMAGE UNDERSTANDING, 2024, 249
  • [34] TOTAL VARIATION BASED PURE QUATERNION DICTIONARY LEARNING METHOD FOR COLOR IMAGE DENOISING
    Wu, Tingting
    Huang, Chaoyan
    Jin, Zhengmeng
    Jia, Zhigang
    Ng, Michael K.
    INTERNATIONAL JOURNAL OF NUMERICAL ANALYSIS AND MODELING, 2022, 19 (05) : 709 - 737
  • [35] A new image segmentation method by minimizing normalized total variation
    Lei, Bohan
    Zhang, Hongwei
    Li, Taihao
    Liu, Shupeng
    Zhang, Min
    Xu, Xiaoyin
    DIGITAL SIGNAL PROCESSING, 2024, 146
  • [36] Local Structural Adaptive Total Variation Method for Image Restoration
    Zeng, Weili
    Lu, Xiaobo
    JOURNAL OF IMAGING SCIENCE AND TECHNOLOGY, 2013, 57 (02)
  • [37] Total Variation Image Restoration Method Based on Subspace Optimization
    Liu, XiaoGuang
    Gao, XingBao
    MATHEMATICAL PROBLEMS IN ENGINEERING, 2018, 2018
  • [38] Nonconvex and nonsmooth total generalized variation model for image restoration
    Zhang, Honglu
    Tang, Liming
    Fang, Zhuang
    Xiang, Changcheng
    Li, Chunyan
    SIGNAL PROCESSING, 2018, 143 : 69 - 85
  • [39] A Nonlocal Total Variation Model for Image Decomposition: Illumination and Reflectance
    Wang, Wei
    Ng, Michael K.
    NUMERICAL MATHEMATICS-THEORY METHODS AND APPLICATIONS, 2014, 7 (03) : 334 - 355
  • [40] An efficient operator splitting method for local region Chan-Vese model
    Hui Wang
    Ting-Zhu Huang
    Jun Liu
    EURASIP Journal on Advances in Signal Processing, 2013