Adaptive Image Denoising Approach Based on Generalized Lp Norm Variational Model

Abstract:

The well-known methods based on gradient dependent regularizers such as total variation (TV) model often suffer the staircase effect and the loss of edge details. In order to overcome such drawbacks, an adaptive variational approach is proposed. First, we introduced a Gaussian smoothed image as the variable of the Lp norm, and then we employed the difference curvature instead of gradient as new edge indicator, which can effectively distinguish between ramps and edges. In the proposed model, the regularization term and fidelity term are both adaptive. At object edges, the regularization term is approximate to the TV norm in order to preserving the edges; in flat and ramp regions, the regularization term is approximate to the L2 norm in order to avoid the staircase effect. Meanwhile, we added a spatially varying fidelity term that locally controls the extent of denoising over image regions according to their content. Local variance measures of the oscillatory part of the signal are to compute the adaptive fidelity term. Comparative results on both natural and medical images demonstrate that the new method can avoid the staircase effect and better preserve fine details than the other variational models.

• Periodical: Applied Mechanics and Materials (Volumes 556-562)
• Pages: 4851-4855
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.556-562.4851
• Online since: May 2014
• Authors: Yi Yan Wang*, Shi Yun Wu
• Keywords: Adaptive Fidelity Term, Difference Curvature, Image De-Noising, Lp Norm, Total Variation

References

[1] L.I. Rudin, S. Osher, E. Fatemi, Nonlinear total variation based noise removal algorithms, Physica D. 1992, (60), p.259–268.

DOI: 10.1016/0167-2789(92)90242-f

[2] S. Osher, M. Burger, D. Goldfarb, J. Xu, W. Yin, An iterative regularization method for total variation based on image restoration, Multi-scale Modeling and Simulation, 2005, 4(2), p.460–489.

DOI: 10.1137/040605412

[3] T. Chan, S. Esedoglu, Aspects of total variation regularized L1 function approximation, SIAM Journal of Applied Mathematics, 2005, 65(5), p.1817–1837.

DOI: 10.1137/040604297

[4] G. Gilboa, Y.Y. Zeevi, N. Sochen, Texture preserving variational denoising usingan adaptive fidelity term, in: Proceedings of the VLSM, Nice, France, 2003, p.137–144.

[5] Song B. Topics in variational PDE image segmentation, inpainting and denoising, Ph.D. Thesis, USA: University of California Los Angeles, June (2003).

[6] T. Chan, A. Marquina, P. Mulet. High-order total variation-based image restoration. SIAM J. SCI. Comput., 2000, vol. 22, no. 2, pp.503-516.

DOI: 10.1137/s1064827598344169

[7] P. Blomgren, T.F. Chan, P. Mulet, Extensions to total variation denoising, in: Proceedings of SPIE, San Diego, vol. 3162, (1997).

[8] P. Blomgren, P. Mulet, T. Chan, C. Wong, Total variation image restoration: numerical methods and extensions, in: ICIP, Santa Barbara, 1997, p.384–387.

DOI: 10.1109/icip.1997.632128

[9] M. Lysaker, X.C. Tai, Iterative image restoration combining total variation minimization and a second-order functional, International Journal of Computer Vision 2006, 66 (1), pp.5-8.

DOI: 10.1007/s11263-005-3219-7

[10] F. Li, C.M. Shen, J.S. Fan, C.L. Shen, Image restoration combing a total variational filter and a fourth-order filter, Journal of Visual Communication and Image Representation , 2007, 18 (4), pp.322-330.

DOI: 10.1016/j.jvcir.2007.04.005

[11] Q. Chen, Q.S. Sen, D.S. Xia, et. al., Adaptive total variation denoising based on difference curvature, Image and Vision Computing, vol. 28, pp.298-306, (2010).

DOI: 10.1016/j.imavis.2009.04.012