Image deblurring models with a mean curvature functional has been widely used to preserve edges and remove the staircase effect in the resulting images. However, the Euler-Lagrange equations of a mean curvature model can be used to solve fourth-order non-linear integro-differential equations. Furthermore, the discretization of fourth-order non-linear integro-differential equations produces an ill-conditioned system so that the numerical schemes like Krylov subspace methods (conjugate gradient etc.) have slow convergence. In this paper, we propose an augmented Lagrangian method for a mean curvature-based primal form of the image deblurring problem. A new circulant preconditioned matrix is introduced to overcome the problem of slow convergence when employing a conjugate gradient method inside of the augmented Lagrangian method. By using the proposed new preconditioner fast convergence has been observed in the numerical results. Moreover, a comparison with the existing numerical methods further reveal the effectiveness of the preconditioned augmented Lagrangian method.
Citation: Shahbaz Ahmad, Faisal Fairag, Adel M. Al-Mahdi, Jamshaid ul Rahman. Preconditioned augmented Lagrangian method for mean curvature image deblurring[J]. AIMS Mathematics, 2022, 7(10): 17989-18009. doi: 10.3934/math.2022991
Image deblurring models with a mean curvature functional has been widely used to preserve edges and remove the staircase effect in the resulting images. However, the Euler-Lagrange equations of a mean curvature model can be used to solve fourth-order non-linear integro-differential equations. Furthermore, the discretization of fourth-order non-linear integro-differential equations produces an ill-conditioned system so that the numerical schemes like Krylov subspace methods (conjugate gradient etc.) have slow convergence. In this paper, we propose an augmented Lagrangian method for a mean curvature-based primal form of the image deblurring problem. A new circulant preconditioned matrix is introduced to overcome the problem of slow convergence when employing a conjugate gradient method inside of the augmented Lagrangian method. By using the proposed new preconditioner fast convergence has been observed in the numerical results. Moreover, a comparison with the existing numerical methods further reveal the effectiveness of the preconditioned augmented Lagrangian method.
[1] | R. Acar, C. R. Vogel, Analysis of bounded variation penalty methods for ill-posed problems, Inverse Probl., 10 (1994), 1217–1229. https://doi.org/10.1088/0266-5611/10/6/003 doi: 10.1088/0266-5611/10/6/003 |
[2] | S. Ahmad, A. M. Al-Mahdi, R. Ahmed, Two new preconditioners for mean curvature-based image deblurring problem, AIMS Math., 6 (2021), 13824–13844. |
[3] | S. Ahmad, F. Fairag, Circulant preconditioners for mean curvature-based image deblurring problem, J. Algorithms Comput., 15 (2021). |
[4] | M. Benzi, G. H. Golub, A preconditioner for generalized saddle point problems, SIAM J. Matrix Anal. A., 26 (2004), 20–41. |
[5] | C. Brito-Loeza, K. Chen, V. Uc-Cetina, Image denoising using the gaussian curvature of the image surface, Numer. Meth. Part. D. E., 32 (2016), 1066–1089. https://doi.org/10.1002/num.22042 doi: 10.1002/num.22042 |
[6] | P. Campisi, K. Egiazarian, Blind image deconvolution: Theory and applications, CRC press, 2016. |
[7] | R. H. Chan, Toeplitz preconditioners for Toeplitz systems with nonnegative generating functions, IMA J. Numer. Anal., 11 (1991), 333–345. |
[8] | R. H. Chan, K. P. Ng, Toeplitz preconditioners for Hermitian Toeplitz systems, Linear Algebra Appl., 190 (1993), 181–208. https://doi.org/10.1016/0024-3795(93)90226-E doi: 10.1016/0024-3795(93)90226-E |
[9] | S. H. Chan, R. Khoshabeh, K. B. Gibson, P. E. Gill, T. Q. Nguyen, An augmented lagrangian method for total variation video restoration, IEEE T. Image Process., 20 (2011), 3097–3111. https://doi.org/10.1109/TIP.2011.2158229 doi: 10.1109/TIP.2011.2158229 |
[10] | T. F. Chan, An optimal circulant preconditioner for Toeplitz systems, SIAM J. Sci. Comput., 9 (1988), 766–771. https://doi.org/10.1137/0909051 doi: 10.1137/0909051 |
[11] | C. Chen, C. Ma, A generalized shift-splitting preconditioner for saddle point problems, Appl. Math. Lett., 43 (2015), 49–55. https://doi.org/10.1016/j.aml.2014.12.001 doi: 10.1016/j.aml.2014.12.001 |
[12] | K. Chen, Introduction to variational image-processing models and applications, Int. J. Comput. Math., 90 (2013), 1–8. |
[13] | K. Chen, F. Fairag, A. Al-Mahdi, Preconditioning techniques for an image deblurring problem, Numer. Linear Algebra, 23 (2016), 570–584. https://doi.org/10.1002/nla.2040 doi: 10.1002/nla.2040 |
[14] | N. R. Choi, A comparative study of non-blind and blind deconvolution of ultrasound images, Dissertations Theses Gradworks, University of Southern California, 2014. |
[15] | F. Di Benedetto, Solution of Toeplitz normal equations by sine transform based preconditioning, Linear Algebra Appl., 285 (1998), 229–255. https://doi.org/10.1016/S0024-3795(98)10115-5 doi: 10.1016/S0024-3795(98)10115-5 |
[16] | F. Fairag, S. Ahmad, A two-level method for image deblurring problem, in 2019 8th International Conference on Modeling Simulation and Applied Optimization (ICMSAO), IEEE, 2019, 1–5. |
[17] | F. Fairag, K. Chen, S. Ahmad, Analysis of the ccfd method for mc-based image denoising problems, Electron. T. Numer. Ana., 54 (2021), 108–127. https://doi.org/10.1553/etna_vol54s108 doi: 10.1553/etna_vol54s108 |
[18] | F. Fairag, K. Chen, C. Brito-Loeza, S. Ahmad, A two-level method for image denoising and image deblurring models using mean curvature regularization, Int. J. Comput. Math., 2021, 1–21. |
[19] | X. Ge, J. Tan, L. Zhang, Y. Qian, Blind image deconvolution via salient edge selection and mean curvature regularization, Signal Process., 190 (2022), 108336. |
[20] | X. M. Gu, Y. L. Zhao, X. L. Zhao, B. Carpentieri, Y. Y. Huang, A note on parallel preconditioning for the all-at-once solution of Riesz fractional diffusion equations, Numer. Math. Theor. Meth. Appl., 14 (2021), 893–919. https://doi.org/10.4208/nmtma.OA-2020-0020 doi: 10.4208/nmtma.OA-2020-0020 |
[21] | L. Guo, X. L. Zhao, X. M. Gu, Y. L. Zhao, Y. B. Zheng, T. Z. Huang, Three-dimensional fractional total variation regularized tensor optimized model for image deblurring, Appl. Math. Comput., 404 (2021), 126224. |
[22] | C. Li, W. Yin, H. Jiang, Y. Zhang, An efficient augmented lagrangian method with applications to total variation minimization, Comput. Optim. Appl., 56 (2013), 507–530. https://doi.org/10.1007/s10589-013-9576-1 doi: 10.1007/s10589-013-9576-1 |
[23] | L. Li, J. Pan, W. S. Lai, C. Gao, N. Sang, M. H. Yang, Learning a discriminative prior for blind image deblurring, IEEE conference on computer vision and pattern recognition, 2018, 6616–6625. |
[24] | F. R. Lin, Preconditioners for block Toeplitz systems based on circulant preconditioners, Numer. Algorithms, 26 (2001), 365–379. https://doi.org/10.1023/A:1016674923507 doi: 10.1023/A:1016674923507 |
[25] | F. R. Lin, W. K. Ching, Inverse Toeplitz preconditioners for Hermitian Toeplitz systems, Numer. Linear Algebra, 12 (2005), 221–229. https://doi.org/10.1002/nla.397 doi: 10.1002/nla.397 |
[26] | F. R. Lin, C. X. Wang, Bttb preconditioners for Bttb systems, Numer. Algorithms, 60 (2012), 153–167. https://doi.org/10.1007/s11075-011-9516-z doi: 10.1007/s11075-011-9516-z |
[27] | M. K. Ng, Iterative methods for Toeplitz systems, London, U.K.: Oxford University Press, 2004. |
[28] | K. L. Riley, Two-level preconditioners for regularized ill-posed problems, PhD Thesis, Montana State University, 1999. |
[29] | L. I. Rudin, S. Osher, E. Fatemi, Nonlinear total variation based noise removal algorithms, Physica D, 60 (1992), 259–268. |
[30] | D. K. Salkuyeh, M. Masoudi, D. Hezari, On the generalized shift-splitting preconditioner for saddle point problems, Appl. Math. Lett., 48 (2015), 55–61. https://doi.org/10.1007/s10986-015-9265-0 doi: 10.1007/s10986-015-9265-0 |
[31] | L. Sun, K. Chen, A new iterative algorithm for mean curvature-based variational image denoising, BIT Numer. Math., 54 (2014), 523–553. |
[32] | X. C. Tai, J. Hahn, G. J. Chung, A fast algorithm for Euler's elastica model using augmented lagrangian method, SIAM J. Imaging Sci., 4 (2011), 313–344. https://doi.org/10.1137/100803730 doi: 10.1137/100803730 |
[33] | X. C. Tai, K. A. Lie, T. F. Chan, S. Osher, Image processing based on partial differential equations: Proceedings of the international conference on PDE-based image processing and related inverse problems, CMA, Oslo, August 8–12, 2005, Springer Science & Business Media, 2006. |
[34] | S. Tao, W. Dong, H. Feng, Z. Xu, Q. Li, Non-blind image deconvolution using natural image gradient prior, Optik, 124 (2013), 6599–6605. https://doi.org/10.1016/j.ijleo.2013.05.068 doi: 10.1016/j.ijleo.2013.05.068 |
[35] | A. N. Tikhonov, Regularization of incorrectly posed problems, Soviet Math. Dokl., 4 (1963), 1624–1627. |
[36] | C. R. Vogel, M. E. Oman, Fast, robust total variation-based reconstruction of noisy, blurred images, IEEE T. Image Process., 7 (1998), 813–824. https://doi.org/10.1109/83.679423 doi: 10.1109/83.679423 |
[37] | C. Wu, X. C. Tai, Augmented lagrangian method, dual methods, and split bregman iteration for rof, vectorial tv, and high order models, SIAM J. Imaging Sci., 3 (2010), 300–339. https://doi.org/10.1137/090767558 doi: 10.1137/090767558 |
[38] | N. Xiong, R. W. Liu, M. Liang, D. Wu, Z. Liu, H. Wu, Effective alternating direction optimization methods for sparsity-constrained blind image deblurring, Sensors, 17 (2017), 174. https://doi.org/10.3390/s17010174 doi: 10.3390/s17010174 |
[39] | F. Yang, K. Chen, B. Yu, D. Fang, A relaxed fixed point method for a mean curvature-based denoising model, Optim. Meth. Softw., 29 (2014), 274–285. |
[40] | J. Zhang, C. Deng, Y. Shi, S. Wang, Y. Zhu, A fast linearised augmented lagrangian method for a mean curvature based model, E. Asian J. Appl. Math., 8 (2018), 463–476. |
[41] | W. Zhu, T. Chan, Image denoising using mean curvature of image surface, SIAM J. Imaging Sci., 5 (2012), 1–32. |
[42] | W. Zhu, X. C. Tai, T. Chan, Augmented lagrangian method for a mean curvature based image denoising model, Inverse Probl. Imag., 7 (2013), 1409–1432. |
[43] | W. Zhu, X. C. Tai, T. Chan, A fast algorithm for a mean curvature based image denoising model using augmented lagrangian method, in Efficient Algorithms for Global Optimization Methods in Computer Vision, Springer, 2014,104–118. |