A Higher Order Regularization Approach for Object Reconstruction with Mixed Laplace-Gaussian Likelihood

doi:10.19596/j.cnki.1001-246x.7831

Abstract

Abstract: A combined first and second order variational model is proposed for reconstructing images corrupted by mixed Laplace-Gaussian noise. The model is constructed by joint maximum a posteriori estimation and expectation maximization. Numerical algorithm is studied by integrating splitting technique into augmented Lagrangian method with modification, such as introduction of adaptively selective functions for preserving details of original images. An adaptive soft-shrinking formulation is advanced for mixed noise removal, in which an alternating minimization algorithm is established. Numerical experiments show validation in tomography reconstruction and image restoration.

Key words: object reconstruction, adaptive soft-shrinking, fourth-order partial differential equation, ADAL

CLC Number:

O241.8

KONG Linghai, KONG Lingbo, XU Haibo, JIA Qinggang. A Higher Order Regularization Approach for Object Reconstruction with Mixed Laplace-Gaussian Likelihood[J]. CHINESE JOURNAL OF COMPUTATIONAL PHYSICS, 2019, 36(3): 280-290.

References

[1] FOI A, TRIMECHE M, EQIAZARIAN K. Practical Poissonian-Gaussian noise modeling and fitting for single-image raw-data[J]. IEEE Trans Image Processing, 2008, 17:1737-1754.
[2] CHOUZENOUX E, JEZIERSKA A, PESQUET J C, et al. A convex approach for image restoration with exact Poisson-Gaussian likelihood[J]. SAIM J Imaging Sci, 2015, 8(4):2662-2682.
[3] CAI J F, CHAN R, NIKOLOVA M. Two-phase methods for deblurring images corrupted by impulse plus Gaussian noise[J]. Inverse Problem Imaging, 2008, 2:187-204.
[4] LIU J, TAI X-C, HUANG H-Y, et al. A weighted dictionary learning model for denoising images corrupted by mixed noise[J]. IEEE Trans Image Processing, 2013,22(3):1108-1120.
[5] GONG Z, SHEN Z W, TOH K C. Image restoration with mixed or unknown[J]. Multiscale Model Simul, 2014, 12(2):458-487.
[6] CALATRONI L, CARLOS DE L R J, SCHÖNLIEB C B. Infimal convolution of data discrepancies for mixed noise removal[J]. SIAM J Imaging Sci, 2017, 10(3):1196-1233.
[7] HANSON K M. Special topics in test methodology:Tomographic reconstruction of axially symmetric objects from a single dynamic radiograph[M]//Progress in Astronautics and Aeronautics, Vol. 155, Washington, 1993:687-698.
[8] BACK N, SCHNEBERK D, MCMILAN C, et al. Research on computed tomography reconstructions from one or two radiographs:A report and the application to FXR radiography[R]. LLNL, 1995,UCLR-ID-120019, 1-58.
[9] PICHOFF N. The new bounds of flash radiography[R], CLEFS CEA, 2006,54:58-66.
[10] THOMPSON M W, HALVERSON D R, WISE G L. Robust detection in nominally Laplace noise[J]. IEEE Trans Commu, 1994, 42(/2/3/4):1651-1660.
[11] BENNING M, BURGER M. Error estimates for general fidelities[J]. Electron Trans Numer Anal, 2011,38:44-68.
[12] RUDIN L, OSHER S, FATEMI E. Nonlinear total variation based noise removal algorithms[J]. Physica D,1992, 60(1-4):259-268.
[13] CHAMBOLLE A, LIONS P. Image recovery via total variation minimization and related problems[J]. Numer Math, 1997,76:167-188.
[14] NIKOLOVA M. A variational approach to remove outliers and impulse noise[J]. J Math Imaging, 2004, 20(1):99-120.
[15] HINTERMVLLER M, LANGER A. Subspace correction methods for a class of nonsmooth and nonadditive convex variational problems with mixed L¹/L² data fidelity in image processing[J]. SIAM J Imaging Sci, 2013, 6(4):2134-2173.
[16] GONG Z, SHEN Z, TOH K C. Image restoration with mixed or unknown[J]. Multiscale Model Simul, 2014, 12(2):458-487.
[17] CHAMBOLLE A, LIONS P L.Image recovery via total variation minimization and related problems[J]. Numer Math, 1997, 76:167-188.
[18] CHAN T, MARQUINA A, MULET P. High-order total variation-based image restoration[J]. SIAM J Sci Comput, 2001,22(2):503-516.
[19] LYSAKER M, LUNDERVOLD A, TAI X C. Noise removal using fourth-order partial differential equation with applications to medical magnetic resonance images in space and time[J]. IEEE Trans Image Process, 2003,12(12):1579-1590.
[20] BERTOZZI A L, GREER J B. Low-curvature image simplifiers:Golobal regularity of smooth solutions and Laplacian limiting schemes[J]. Commun Pure Appl Math, 2004, 57(6):764-790.
[21] LYSAKER M, TAI X C. Iterative image restoration combining total variation minimization and a second-order functional[J]. Int J Comput Vis, 2006, 66(1):5-18.
[22] LI F, SHEN C M, FAN J, et al. Image restoration combining a total variational filter and a fourth-order filter[J]. J Vis Commun Image R, 2007, 18:322-330.
[23] BREDIES K, KUNISCH K, POCK T. Total generalized variation[J]. SIAM J Imaging Sci, 2010,3(3):1-42.
[24] WU T T, YANG Y F, PANG Z F. A modified fixed-point iterative algorithm for image restoration using fourth-order PDE model[J]. Applied Numer Math, 2012, 62:79-90.
[25] BENNING M, BRUNE C, BURGER M, et al. Higher-order TV methods-enhancement via Bregman iteration[J]. J Sci Comput, 2013, 54(2-3):269-310.
[26] PAPAFITSOROS K, SHCÖNLIEB C B. A combined first and second order variational approach for image reconstruction[J]. Journal of Math Imaging Vis, 2014, 48:308-338.
[27] STEIDL G. Combined first and second order variational approaches for image processing[J]. Jahresber Dtsch Math-Ver, 2015, 177:133-160.
[28] ASAKI T J, CHARTRAND R, VIXIE K R, WOHLBERG B. Abel inversion using total-variation regularization[J]. Inverse Problems, 2005, 21:1895-1903.
[29] ABRAHAM R, BERGOUNIOUX M, TRÉLAT E. A penalization approach for tomographic reconstruction of binary axially symmetric objects[J]. Appl Math Optim, 2008, 58:345-371.
[30] CHAN R H, LIANG H, WEI S H, et al. High-order total variation regularization approach for axially symmetric object tomography from a single radiograph[J]. Inverse Problems & Imaging, 2015, 9(1):55-77.
[31] IDIER J. Bayesian approach to inverse problems[M]. New York:Wiley, 2008.
[32] BURGER B, MENNUCCI A C G, OSHER S, et al. Level set and PDE based reconstruction methods in imaging[M]. New York:Springer, 2008:1-70.
[33] DEMENGEL F. Fornctions a hessian borne[J]. Ann Inst Fourier, 1985, 34:155-190.
[34] LIU J, HUANG H Y, HUAN Z D, et al. Adaptive variational method for restoring color images with high density impulse noise[J]. Inter J Comput Vis, 2010, 90:131-149.
[35] LIU J, ZHANG H L. Image segmentation using a local GMM in a variational framework[J]. J Math Imaging Vis, 2013,46:161-176.
[36] GABAY D. Application of the method of multipliers to variational inequalites[M]//Fortin M, Glowinski R. Augmented Lagrange method:Applications to the solutions of boundary valued problems. 1983, North Holland, Amsterdam:299-331.
[37] BERTSEKAS D P, TSITSIKLIS J N. Parallel and distributed computation:Numerical methods[M]. Belmont Massachusetts:Athena Scientific, 1997.
[38] SETZER S. Operator splittings, Bregman methods and frame shrinkage in image processing[J]. Int J Cmput Vis, 2011, 92(3):265-280.
[39] GOPI V P, PALANISAMY P, WAHID K A, et al. Micro-CT image reconstruction based on alternating direction augmented Lagrangian method and total variation[J]. Computerized Medical Imaging and Graphics, 2013,37:419-429.
[40] BURGER M, SAWATZKY A, STEIDL G. First order algorithm in variational image processing[EB/OL]. 2014, arXiv:1412, 4237.
[41] OSHER S, PARAGIOS N. Geometric level set methods in imaging, vision, and graphics[M]. New York:Springer-Verlag, 2003.
[42] XU H, WEI S. A conjugate gradient algorithm for density reconstruction in high-energy X-ray radiography[J]. Chinese Journal of Computational Physice, 2006,23(2):144-150.