Multimedia Tools and Applications

, Volume 77, Issue 22, pp 30121–30134 | Cite as

Selection of regularization parameter in GMM based image denoising method

  • Yuhui Zheng
  • Min Li
  • Jianwei Zhang
  • Jin WangEmail author


Currently, the image denoising methods using Gaussian mixture model to learn image prior have received much attention. Among these methods, expected patch log likelihood based image denoising approach has been shown to be surprisingly competitive in image restoration. However, recent related works generally utilize global regularization parameter that influences the performance of denoising algorithm. In this paper, with the consideration that the Gaussian mixture model has the capability of clustering, we propose an adaptive estimation method of regularization parameter for expected patch log likelihood based image denoising. Our method jointly employs the Lagrange multiplier technique and entropy concept to select regularization parameter for each underlying cluster. Experimental results illustrate the relatively good performance of our image denoising method in terms of visual improvement and peak signal to noise ratio.


Image denoising Gaussian mixture model Regularization parameter selection Lagrange multiplier method 



This work was supported by the National Natural Science Foundation of China under Grants 61572257 and 61672295, the Natural Science Fund for Colleges and Universities in Jiangsu Province (15KJB520025), and the PAPD (a project funded by the priority academic program development of Jiangsu Higher Education Institutions).


  1. 1.
    Buades A, Coll B, Morel J, Sbert C (2009) Self-similarity driven color demosaicking. IEEE Trans Image Process 18(6):1192–1202MathSciNetCrossRefGoogle Scholar
  2. 2.
    Chen K, Piccolomini E-L, Zama F (2014) An automatic regularization parameter selection algorithm in the total variation model for image deblurring. Numer Algorithms 67(1):73–92MathSciNetCrossRefGoogle Scholar
  3. 3.
    Dong Y, Hintermüller M, Camacho M-R (2011) Automated regularization parameter selection in a multi-scale total variation model for image restoration. J Math Imaging Vision 40(1):82–104MathSciNetCrossRefGoogle Scholar
  4. 4.
    Elad M, Aharon M (2006) Image denoising via sparse and redundant representations over learned dictionaries. IEEE Trans Image Process 15(12):3736–3745MathSciNetCrossRefGoogle Scholar
  5. 5.
    Gilboa G, Sochen N, Zeevi YY (2006) Variational denoising of partly textured images by spatially varying constraint. IEEE Trans Image Process 15(8):2281–2289CrossRefGoogle Scholar
  6. 6.
    Gonzalez D-S, Moreno A-J, Enriquez E-M, Maria F (2014) Improved method to select the Lagrange multiplier for rate-distortion based motion estimation in video coding. IEEE Trans Circuits Syst Video Technol 24(3):452–464CrossRefGoogle Scholar
  7. 7.
    Han J, Quan R, Zhang D (2018) Robust object co-segmentation using background prior. IEEE Trans Image Process 27(4):1639–1651MathSciNetCrossRefGoogle Scholar
  8. 8.
    Izmailov A, Uskov F (2015) Attraction of newton method to critical Lagrange multipliers: fully quadratic case. Math Program 152(1-2):33–73MathSciNetCrossRefGoogle Scholar
  9. 9.
    Jeong S, Lee Y, Lee S (2017) Development of an automatic sorting system for fresh ginsengs by image processing techniques. Hum-cent comput info 7:41. CrossRefGoogle Scholar
  10. 10.
    Koo K, Cha E (2017) Image recognition performance enhancements using image normalization. Hum-cent comput info 7:33. CrossRefGoogle Scholar
  11. 11.
    Lee I, Moon B (2017) An improved stereo matching algorithm with robustness to noise based on adaptive support weight. J Inf Process Syst 13(2):256–267Google Scholar
  12. 12.
    Lou Y, Zhang X, Osher S, Bertozzi A (2010) Image recovery via nonlocal operators. J Sci Comput 42(2):185–197MathSciNetCrossRefGoogle Scholar
  13. 13.
    Lu X, Lin Z, Jin H (2015) Image-specific prior adaption for denoising. IEEE Trans Image Process 24(12):5469–5478MathSciNetCrossRefGoogle Scholar
  14. 14.
    Niknejad M, Rabbani H, Massound B-Z (2015) Image restoration using Gaussian mixture models with spatially constrained patch clustering. IEEE Trans Image Process 24(11):3624–3636MathSciNetCrossRefGoogle Scholar
  15. 15.
    Pan Z, Lei J, Zhang Y, Sun X, Kwong S (2016) Fast motion estimation based on content property for low-complexity H.265/HEVC encoder. IEEE Trans Broadcast 62(3):675–684CrossRefGoogle Scholar
  16. 16.
    Papyan V, Elad M (2016) Multi-scale patch-based image restoration. IEEE Trans Image Process 25(1):249–261MathSciNetCrossRefGoogle Scholar
  17. 17.
    Park J (2017) Efficient approaches to computer vision and pattern recognition. J Inf Process Syst 13(6):1431–1435Google Scholar
  18. 18.
    Ren J, Liu J, Guo Z (2013) Context-aware sparse decomposition for image denoising and super-resolution. IEEE Trans Image Process 22(4):1456–1469MathSciNetCrossRefGoogle Scholar
  19. 19.
    Rezghi M, Hosseini S-M (2009) A new variant of L-curve for Tikhonov regularization. J Comput Appl Math 231(2):914–924MathSciNetCrossRefGoogle Scholar
  20. 20.
    Rudin L, Osher S, Fatemi E (1992) Nonlinear total variation based noise removal algorithms. Physica D 60(1-4):259–268MathSciNetCrossRefGoogle Scholar
  21. 21.
    Su Z, Yang L, Zhu S, Si N, Lv X (2017) Gaussian mixture image restoration based on maximum correntropy criterion. Electron Lett 53(11):715–716CrossRefGoogle Scholar
  22. 22.
    Wen Y, Chan R (2012) Parameter selection for total-variation-based image restoration using discrepancy principle. IEEE Trans Image Process 21(4):1770–1781MathSciNetCrossRefGoogle Scholar
  23. 23.
    Xiao F, Liu W, Li Z, Chen L (2018) Noise-tolerant wireless sensor networks localization via multi-norms regularized matrix completion. IEEE Trans Veh Technol 67(3):2409–2419CrossRefGoogle Scholar
  24. 24.
    Yan R, Ling S, Liu Y (2013) Nonlocal hierarchical dictionary learning using wavelets for image denoising. IEEE Trans Image Process 22(12):4689–4698MathSciNetCrossRefGoogle Scholar
  25. 25.
    Yang Z, Jacob M (2013) Nonlocal regularization of inverse problems: a unified variational framework. IEEE Trans Image Process 22(8):3192–3203CrossRefGoogle Scholar
  26. 26.
    Yao X, Han J, Zhang D, Nie F (2017) Revisiting co-saliency detection: a novel approach based on two-stage multi-view spectral rotation co-clustering. IEEE Trans Image Process 26(7):3196–3209MathSciNetCrossRefGoogle Scholar
  27. 27.
    Yuan Q, Zhang L, Shen H, Li P (2010) Adaptive multiple-frame image super-resolution based on U-curve. IEEE Trans Image Process 19(12):3157–3170MathSciNetCrossRefGoogle Scholar
  28. 28.
    Zeng Y-H, Peng Z, Yang Y-F (2016) A hybrid splitting method for smoothing Tikhonov regularization problem. J Inequal Appl 1:1–13MathSciNetzbMATHGoogle Scholar
  29. 29.
    Zhang J, Zhao D, Gao W (2014) Group-based sparse representation for image restoration. IEEE Trans Image Process 23(8):3336–3351MathSciNetCrossRefGoogle Scholar
  30. 30.
    Zhang J, Yu Q, Zheng Y, Zhang H, Wu J (2016) Regularization parameter selection for TV image denoising using spatially adaptive local spectral response. J Internet technol 17(6):1117–1124Google Scholar
  31. 31.
    Zhang J, Liu J, Li T, Zheng Y, Wang J (2017) Gaussian mixture model learning based image denoising method with adaptive regularization parameters. Multimed Tools Appl 76(9):11471–11483CrossRefGoogle Scholar
  32. 32.
    Zheng Y, Jeon B, Zhang J, Chen Y (2015) Adaptively determining regularization parameters in non-local total variation regularization for image denoising. Electron Lett 51(2):144–145CrossRefGoogle Scholar
  33. 33.
    Zheng Y, Zhou X, Jeon B, Shen J, Zhang H (2017) Multi-scale patch prior learning for image denoising using Student’s-t mixture model. J Internet technol 18(7):1553–1560Google Scholar
  34. 34.
    Zheng Y, Jeon B, Sun L, Zhang J, Zhang H (2017) Student's t-hidden Markov model for unsupervised learning using localized feature selection. IEEE Trans Circuits Syst Video Technol.
  35. 35.
    Zheng Y, Ma K, Yu Q, Zhang J, Wang J (2017) Regularization parameter selection for total variation model based on local spectral response. J Inf Process Syst 13(5):1168–1182Google Scholar
  36. 36.
    Zoran D, Weiss Y (2011) From learning models of natural image patches to whole image restoration. International Conference on Computer Vision, p 479–486Google Scholar

Copyright information

© Springer Science+Business Media, LLC, part of Springer Nature 2018

Authors and Affiliations

  1. 1.College of Computer and SoftwareNanjing University of Information Science & TechnologyNanjingChina
  2. 2.College of Math and StatisticsNanjing University of Information Science & TechnologyNanjingChina
  3. 3.College of Information EngineeringYangzhou UniversityYangzhouChina

Personalised recommendations