A convex-nonconvex variational method for the additive decomposition of functions on surfaces. (English) Zbl 1468.94016
Summary: We present a convex-nonconvex variational approach for the additive decomposition of noisy scalar fields defined over triangulated surfaces into piecewise constant and smooth components. The energy functional to be minimized is defined by the weighted sum of three terms, namely an \(\ell_2\) fidelity term for the noise component, a Tikhonov regularization term for the smooth component and a total variation (TV)-like non-convex term for the piecewise constant component. The last term is parametrized such that the free scalar parameter allows to tune its degree of non-convexity and, hence, to separate the piecewise constant component more effectively than by using a classical convex TV regularizer without renouncing to convexity of the total energy functional. A method is also presented for selecting the two regularization parameters. The unique solution of the proposed variational model is determined by means of an efficient ADMM-based minimization algorithm. Numerical experiments show a nearly perfect separation of the different components.
MSC:
| 94A08 | Image processing (compression, reconstruction, etc.) in information and communication theory |
| 94A12 | Signal theory (characterization, reconstruction, filtering, etc.) |
| 65K10 | Numerical optimization and variational techniques |
Keywords:
variational image decomposition; functions on surfaces; convex non-convex strategy; convex non-convex optimization; image decomposition; surface processingReferences:
| [1] | Aujol J F, Gilboa G, Chan T and Osher S 2006 Structure-texture image decomposition modeling, algorithms, and parameter selection Int. J. Comput. Vis.67 111-36 · Zbl 1287.94011 · doi:10.1007/s11263-006-4331-z |
| [2] | Chambolle A and Lions P L 1997 Image recovery via total variation minimization and related problems Numer. Math.76 167-88 · Zbl 0874.68299 · doi:10.1007/s002110050258 |
| [3] | Osher S, Solé A and Vese L 2003 Image decomposition and restoration using total variation minimization and the H1 norm Multiscale Model. Simul.1 349-70 · Zbl 1051.49026 · doi:10.1137/S1540345902416247 |
| [4] | Vese L and Osher S 2003 Modeling textures with total variation minimization and oscillating patterns in image processing J. Sci. Comput.19 553-72 · Zbl 1034.49039 · doi:10.1023/A:1025384832106 |
| [5] | Gholami A and Hosseini S M 2013 A balanced combination of Tikhonov and total variation regularizations for reconstruction of piecewise-smooth signals Signal Process.93 1945-60 · doi:10.1016/j.sigpro.2012.12.008 |
| [6] | Meyer Y 2001 Oscillating Patterns in Image Processing and Nonlinear Evolution Equations: the 15th Dean Jacqueline B Lewis Memorial Lectures(University Lecture Series vol 22) (Providence, RI: American Mathematical Society) · Zbl 0987.35003 · doi:10.1090/ulect/022 |
| [7] | Blake A and Zisserman A 1987 Visual Reconstruction (Cambridge, MA: MIT Press) · doi:10.7551/mitpress/7132.001.0001 |
| [8] | Lanza A, Morigi S, Sgallari F and Yezzi A J 2013 Variational image denoising based on autocorrelation whiteness SIAM J. Imaging Sci.6 1931-55 · Zbl 1279.68330 · doi:10.1137/120885504 |
| [9] | Chen P Y and Selesnick I W 2014 Group-sparse signal denoising: non-convex regularization, convex optimization IEEE Trans. Signal Process.62 3464-78 · Zbl 1394.94115 · doi:10.1109/TSP.2014.2329274 |
| [10] | Lanza A, Morigi S and Sgallari F 2015 Convex image denoising via non-convex regularization Scale Space Variational Methods Comput. Vis.9087 666-77 · Zbl 1444.94015 · doi:10.1007/978-3-319-18461-6_53 |
| [11] | Lanza A, Morigi S and Sgallari F 2016 Convex image denoising via non-convex regularization with parameter selection J. Math. Imaging Vis.56 195-220 · Zbl 1391.94088 · doi:10.1007/s10851-016-0655-7 |
| [12] | Lanza A, Morigi S, Selesnick I and Sgallari F 2017 Nonconvex nonsmooth optimization via convex-nonconvex majorization-minimization Numer. Math.136 343-81 · Zbl 1368.65087 · doi:10.1007/s00211-016-0842-x |
| [13] | Nikolova M 1998 Estimation of binary images by minimizing convex criteria Proc. IEEE Int. Conf. Image Process.2 108-12 · doi:10.1109/ICIP.1998.723327 |
| [14] | Parekh A and Selesnick I W 2015 Convex denoising using non-convex tight frame regularization IEEE Signal Processing Letters22 1786-90 · doi:10.1109/LSP.2015.2432095 |
| [15] | Rudin L I, Osher S and Fatemi E 1992 Nonlinear total variation based noise removal algorithms Physics D 60 259-68 · Zbl 0780.49028 · doi:10.1016/0167-2789(92)90242-F |
| [16] | Selesnick I W and Bayram I 2014 Sparse signal estimation by maximally sparse convex optimization IEEE Trans. Signal Process.62 1078-92 · Zbl 1394.94510 · doi:10.1109/TSP.2014.2298839 |
| [17] | Selesnick I W, Parekh A and Bayram I 2015 Convex 1D total variation denoising with non-convex regularization IEEE Signal Process. Lett.22 141-4 · doi:10.1109/LSP.2014.2349356 |
| [18] | Botsch M, Kobbelt L, Pauly M, Alliez P and Levy B 2010 Polygon Mesh Processing ed A K Peters (Boca Raton, FL: CRC Press) · doi:10.1201/b10688 |
| [19] | Gilboa G 2014 A total variation spectral framework for scale and texture analysis SIAM J. Imaging Sci.7 1937-61 · Zbl 1361.94014 · doi:10.1137/130930704 |
| [20] | Horesh D and Gilboa G 2016 Separation surfaces in the spectral TV domain for texture decomposition IEEE Trans. Image Process.25 4260-70 · Zbl 1408.94251 |
| [21] | Zhang C-H 2010 Nearly unbiased variable selection under minimax concave penalty Ann. Stat.38 894-942 · Zbl 1183.62120 · doi:10.1214/09-AOS729 |
| [22] | Chan R, Lanza A, Morigi S and Sgallari F 2018 Convex non-convex image segmentation Numer. Math.138 635-80 · Zbl 1453.65141 · doi:10.1007/s00211-017-0916-4 |
| [23] | Tsitsas N L 2010 On block matrices associated with discrete trigonometric transforms and their use in the theory of wave propagation J. Comput. Math.6 864-78 · Zbl 1240.65398 · doi:10.4208/jcm.1004-m3193 |
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.
