×

Superresolution from a single noisy image by the median filter transform. (English) Zbl 1333.94012

Summary: The task of single-image superresolution is to obtain a high resolution image from a low resolution input. Many strategies have achieved considerable success by learning from image databases so as to find suitable replacements for the missing information. However, little effort has been devoted to coping with significant amounts of noise in the input. Noise makes the problem even harder, since its distribution is unknown in many practical cases. Here an approach is proposed to solve this problem irrespective of the noise type, which supports both integer and fractional zoom factors. It is based on the application of the median filter on parallelogram shaped windows chosen according to a suitable probability distribution. The resulting outputs are then median filtered again to obtain the final output. Experimental results are reported for artificial and natural images under Gaussian and impulsive uniform noise. It is found that the proposed approach outperforms several state-of-the-art single-image superresolution methods both quantitatively and qualitatively.

MSC:

94A08 Image processing (compression, reconstruction, etc.) in information and communication theory
68U10 Computing methodologies for image processing
62G30 Order statistics; empirical distribution functions
62E20 Asymptotic distribution theory in statistics
PDFBibTeX XMLCite
Full Text: DOI

References:

[1] H. A. Aly and E. Dubois, {\it Image up-sampling using total-variation regularization with a new observation model}, IEEE Trans. Image Process., 14 (2005), pp. 1647-1659.
[2] G. W. Basset, {\it Equivariant, monotonic, 50}
[3] M. Z. A. Bhotto, M. O. Ahmad, and M. N. S. Swamy, {\it Fine granularity spatially adaptive regularization for TVL \(1\) based image deblurring}, in Proceedings of the 27th IEEE Canadian Conference on Electrical and Computer Engineering (CCECE), IEEE, Washington, DC, 2014, pp. 1-5.
[4] N. K. Bose and N. A. Ahuja, {\it Superresolution and noise filtering using moving least squares}, IEEE Trans. Image Process., 15 (2006), pp. 2239-2248.
[5] A. Buades, B. Coll, and J.-M. Morel, {\it A non-local algorithm for image denoising}, in Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR, Vol. 2, IEEE, Washington, DC, 2005, pp. 60-65. · Zbl 1108.94004
[6] J. T. Chu, {\it On the distribution of the sample median}, Ann. Math. Statist., 26 (1955), pp. 112-116. · Zbl 0064.13102
[7] H. A. David and H. N. Nagaraja, {\it Order Statistics}, 3rd ed., Wiley, Hoboken, NJ, 2003. · Zbl 1053.62060
[8] W. Dong, D. Zhang, G. Shi, and X. Wu, {\it Image deblurring and super-resolution by adaptive sparse domain selection and adaptive regularization}, IEEE Trans. Image Process., 20 (2011), pp. 1838-1857. · Zbl 1372.94072
[9] W. Dong, L. Zhang, R. Lukac, and G. Shi, {\it Sparse representation based image interpolation with nonlocal autoregressive modeling}, IEEE Trans. Image Process., 22 (2013), pp. 1382-1394. · Zbl 1373.94103
[10] W. Dong, L. Zhang, G. Shi, and X. Li, {\it Nonlocally centralized sparse representation for image restoration}, IEEE Trans. Image Process., 22 (2013), pp. 1620-1630. · Zbl 1373.94104
[11] V. Duval, {\it Variational and Non-local Methods in Image Processing: A Geometric Study}, Ph.D. thesis, Telecom ParisTech, \burlhttps://who.rocq.inria.fr/Vincent.Duval/files/these_fr.pdf, 2013.
[12] V. Duval, J.-F. Aujol, and Y. Gousseau, {\it The TVL1 model: A geometric point of view}, Multiscale Model. Simul., 8 (2009), pp. 154-189. · Zbl 1187.94010
[13] Y. C. Eldar and T. Michaeli, {\it Beyond bandlimited sampling: A review of nonlinearities, smoothness, and sparsity}, IEEE Signal Process. Mag., 26 (2009), pp. 48-68.
[14] S. Farsiu, D. Robinson, M. Elad, and P. Milanfar, {\it Advances and challenges in super-resolution}, Internat. J. Imaging Syst. Tech., 14 (2004), pp. 47-57.
[15] W. Feller, {\it An Introduction to Probability Theory and Its Applications, Vol}. 1, 3rd ed., Wiley, New York, 1968, Ch. Laws of Large Numbers, pp. 228-247.
[16] A. Foi, M. Trimeche, V. Katkovnik, and K. Egiazarian, {\it Practical Poissonian-Gaussian noise modeling and fitting for single-image raw-data}, IEEE Trans. Image Process., 17 (2008), pp. 1737-1754. · Zbl 1371.94492
[17] G. Freedman and R. Fattal, {\it Image and video upscaling from local self-examples}, ACM Trans. Graphics, 30 (2011), 12.
[18] Z. Gan, L. Qi, C. Chen, Z. Cui, and X. Zhu, {\it Novel reconstruction based image super resolution}, J. Comput. Inform. Syst., 9 (2013), pp. 8913-8920.
[19] D. Glasner, S. Bagon, and M. Irani, {\it Super-resolution from a single image}, in Proceedings of the 12th IEEE International Conference on Computer Vision, IEEE, Washington, DC, 2009, pp. 349-356.
[20] D. Glasner, S. Bagon, and M. Irani, {\it Super-resolution from a Single Image}, available online from http://www.wisdom.weizmann.ac.il/ vision/SingleImageSR.html, 2009.
[21] F. Guichard, J. M. Morel, and R. Ryan, {\it Contrast Invariant Image Analysis and PDE’s}, Tech. report, IPOL Development Server, http://dev.ipol.im/ morel/JMMBookOct04.pdf, 2004.
[22] H. He and W.-C. Siu, {\it Single image super-resolution using Gaussian process regression}, in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2011, IEEE, Washington, DC, pp. 449-456.
[23] J. S. Huang, {\it Estimating the variance of the sample median, discrete case}, Statist. Probab. Lett., 11 (1991), pp. 291-298. · Zbl 0722.62031
[24] P. J. Huber and E. M. Ronchetti, {\it Robust Statistics}, 2nd ed., Wiley, Hoboken, NJ, 2009. · Zbl 1276.62022
[25] H. Jeffreys, {\it An invariant form for the prior probability in estimation problems}, Proc. Roy. Soc. London Ser. A, 186 (1946), pp. 453-461. · Zbl 0063.03050
[26] R. Keys, {\it Cubic convolution interpolation for digital image processing}, IEEE Trans. Acoustics Speech Signal Process., 29 (1981), pp. 1153-1160. · Zbl 0524.65006
[27] D.-O. Kim and R.-H. Park, {\it Image quality assessment using the amplitude/phase quantization code}, IEEE Trans. Consumer Electronics, 56 (2010), pp. 2756-2762.
[28] K. I. Kim and Y. Kwon, {\it Single-image super-resolution using sparse regression and natural image prior}, IEEE Trans. Pattern Anal. Machine Intell., 32 (2010), pp. 1127-1133.
[29] X. Li and M. T. Orchard, {\it New edge-directed interpolation}, IEEE Trans. Image Process., 10 (2001), pp. 1521-1527.
[30] G. D. Lin and J. S. Huang, {\it Variances of sample medians}, Statist. Probab. Lett., 8 (1989), pp. 143 – 146. · Zbl 0676.62024
[31] Z. Lin and H.-Y. Shum, {\it Fundamental limits of reconstruction-based superresolution algorithms under local translation}, IEEE Trans. Pattern Anal. Machine Intell., 26 (2004), pp. 83-97.
[32] I. Lissner, J. Preiss, P. Urban, M. S. Lichtenauer, and P. Zolliker, {\it Image-difference prediction: From grayscale to color}, IEEE Trans. Image Process., 22 (2013), pp. 435-446. · Zbl 1373.94254
[33] E. López-Rubio, {\it A histogram transform for probability density function estimation}, IEEE Trans. Pattern Anal. Machine Intell., 36 (2014), pp. 644-656.
[34] Y. Ma, M. G. Genton, and E. Parzen, {\it Asymptotic properties of sample quantiles of discrete distributions}, Ann. Inst. Statist. Math., 63 (2011), pp. 227-243. · Zbl 1432.62035
[35] S. Mallat and G. Yu, {\it Super-resolution with sparse mixing estimators}, IEEE Trans. Image Process., 19 (2010), pp. 2889-2900. · Zbl 1371.94252
[36] G. Meyer, {\it The University of Waterloo Image Repository}, \burlhttp://links.uwaterloo.ca/Repository.html, 2014.
[37] F. Mezzadri, {\it How to generate random matrices from the classical compact groups}, Notices Amer. Math. Soc., 54 (2007), pp. 592-604. · Zbl 1156.22004
[38] C. H. Ooi and N. A. M. Isa, {\it Quadrants dynamic histogram equalization for contrast enhancement}, IEEE Trans. Consumer Electron., 56 (2010), pp. 2552-2559.
[39] T. Peleg and M. Elad, {\it A statistical prediction model based on sparse representations for single image super-resolution}, IEEE Trans. Image Process., 23 (2014), pp. 2569-2582. · Zbl 1374.94299
[40] J. Portilla, V. Strela, M. J. Wainwright, and E. P. Simoncelli, {\it Image denoising using scale mixtures of Gaussians in the wavelet domain}, IEEE Trans. Image Process., 12 (2003), pp. 1338-1351. · Zbl 1279.94028
[41] F. Qiu, Y. Xu, C. Wang, and Y. Yang, {\it Noisy image super-resolution with sparse mixing estimators}, in Proceedings of the 4th International Congress on Image and Signal Processing (CISP), Vol. 2, IEEE, Washington, DC, 2011, pp. 1081-1085.
[42] K. N. Ramamurthy, J. J. Thiagarajan, and A. Spanias, {\it Recovering non-negative and combined sparse representations}, Digit. Signal Process., 26 (2014), pp. 21-35.
[43] P. J. Rousseeuw and G. W. Bassett, {\it The remedian: A robust averaging method for large data sets}, J. Amer. Statist. Assoc., 85 (1990), pp. 97-104. · Zbl 0702.62030
[44] Y. Shen, B. Han, and E. Braverman, {\it Removal of mixed Gaussian and impulse noise using directional tensor product complex tight framelets}, J. Math. Imaging Vision, to appear. · Zbl 1339.94012
[45] A. Singh and N. Ahuja, {\it Sub-band energy constraints for self-similarity based super-resolution}, in Proceedings of the 22nd International Conference on Pattern Recognition, Stockholm, Sweden, 2014, pp. 4447-4452.
[46] A. Singh and N. Ahuja, {\it Super-resolution using sub-band self-similarity}, in Proceedings of the 12th Asian Conference on Computer Vision, Vol. II, Singapore, 2014, pp. 552-568.
[47] A. Singh, F. Porikli, and N. Ahuja, {\it Super-resolving noisy images}, in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), IEEE, Washington, DC, 2014, pp. 2846-2853.
[48] J. Sun, J. Zhu, and M. F. Tappen, {\it Context-constrained hallucination for image super-resolution}, in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), IEEE, Washington, DC, 2010, pp. 231-238.
[49] J. Suo, Y. Deng, L. Bian, and Q. Dai, {\it Joint non-Gaussian denoising and superresolving of raw high frame rate videos}, IEEE Trans. Image Process., 23 (2014), pp. 1154-1168. · Zbl 1374.94361
[50] H. Takeda, {\it Kernel Regression-Based Image Processing Toolbox}, \burlhttp://alumni.soe.ucsc.edu/ htakeda/KernelToolBox.html, 2014.
[51] J. Tian and K.-K. Ma, {\it A survey on super-resolution imaging}, Signal Image Video Process., 5 (2011), pp. 329-342.
[52] R. Timofte, V. De, and L. Van Gool, {\it Anchored neighborhood regression for fast example-based super-resolution}, in Proceedings of the IEEE International Conference on Computer Vision, ICCV ’13, IEEE Computer Society, Washington, DC, 2013, pp. 1920-1927.
[53] R. Timofte, V. De Smet, and L. Van Gool, {\it Adjusted anchored neighborhood regression for fast super-resolution}, in Proceedings of the 12th Asian Conference on Computer Vision, Part IV, Singapore, 2014, pp. 111-126.
[54] Y. Traonmilin, S. Ladjal, and A. Almansa, {\it Outlier removal power of the L \(1\)-norm super-resolution}, in Scale Space and Variational Methods in Computer Vision, A. Kuijper, K. Bredies, T. Pock, and H. Bischof, eds., Lecture Notes in Comput. Sci. 7893, Springer, Berlin, 2013, pp. 198-209.
[55] Y. Traonmilin, S. Ladjal, and A. Almansa, {\it Robust multi-image processing with optimal sparse regularization}, J. Math. Imaging Vision, 51 (2015), pp. 413-429. · Zbl 1330.68333
[56] J. W. Tukey, {\it Contributions to survey sampling and applied statistics in honor of H.O. Hartley}, Academic Press, New York, 1978, Ch. The Ninther, a Technique for Low-Effort Robust (Resistant) Location in Large Samples, pp. 251-257.
[57] Q. Wang and R. K. Ward, {\it A new orientation-adaptive interpolation method}, IEEE Trans. Image Process., 16 (2007), pp. 889-900.
[58] Z. Wang, A. C. Bovik, H. R. Sheikh, and E. P. Simoncelli, {\it Image quality assessment: From error visibility to structural similarity}, IEEE Trans. Image Process., 13 (2004), pp. 600-612.
[59] A. Weber, {\it USC-SIPI Image Database}, \burlhttp://sipi.usc.edu/database/, 2014.
[60] W. Xie, J. Liu, L. Shao, and F. Jing, {\it Multi-scale fractal coding for single image super-resolution}, in Intelligent Computing Theory, D.-S. Huang, V. Bevilacqua, and P. Premaratne, eds., Lecture Notes in Comput. Sci. 8588, Springer, Berlin, 2014, pp. 425-434.
[61] C.-Y. Yang, C. Ma, and M.-H. Yang, {\it Single-image super-resolution: A benchmark}, in Proceedings of European Conference on Computer Vision (ECCV), Part IV, Zurich, Switzerland, 2014, pp. 372-386.
[62] C.-Y. Yang and M.-H. Yang, {\it Fast direct super-resolution by simple functions}, in Proceedings of the IEEE International Conference on Computer Vision (ICCV), IEEE, Washington, DC, 2013, pp. 561-568.
[63] J. Yang, J. Wright, T. S. Huang, and Y. Ma, {\it Image super-resolution via sparse representation}, IEEE Trans. Image Process., 19 (2010), pp. 2861-2873. · Zbl 1371.94429
[64] W. Yin, D. Goldfarb, and S. Osher, {\it The total variation regularized \(L^1\) model for multiscale decomposition}, Multiscale Model. Simul., 6 (2007), pp. 190-211. · Zbl 1355.49037
[65] G. Yu, G. Sapiro, and S. Mallat, {\it Solving inverse problems with piecewise linear estimators: From Gaussian mixture models to structured sparsity}, IEEE Trans. Image Process., 21 (2012), pp. 2481-2499. · Zbl 1373.94471
[66] W.-S. Yu, M.-H. Wang, and Z.-J. Huang, {\it An adaptively edge-guided back-projection algorithm for single image super-resolution}, Internat. J. Adv. Comput. Tech., 4 (2012), pp. 232-239.
[67] K. Zhang, X. Gao, D. Tao, and X. Li, {\it Single image super-resolution with non-local means and steering kernel regression}, IEEE Trans. Image Process., 21 (2012), pp. 4544-4556. · Zbl 1373.94482
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.