Local sparse representation for astronomical image denoising

Abstract

Motivated by local coordinate coding (LCC) theory in nonlinear manifold learning, a new image representation model called local sparse representation (LSR) for astronomical image denoising was proposed. Borrowing ideas from surrogate function and applying the iterative shrinkage-thresholding algorithm (ISTA), an iterative shrinkage operator for LSR was derived. Meanwhile, a fast approximated LSR method by first performing a K-nearest-neighbor search and then solving a l¹ optimization problem was presented under the guarantee of denoising performance. In addition, the LSR model and adaptive dictionary learning were incorporated into a unified optimization framework, which explicitly established the inner connection of them. Such processing allows us to simultaneously update sparse coding vectors and the dictionary by alternating-optimization method. The experimental results show that the proposed method is superior to the traditional denoising method and reaches state-of-the-art performance on astronomical image.

Keywords

astronomical image denoising / local sparse representation (LSR) / dictionary learning / alternating optimization

References

[1] HarmelingS, HirschM, SraS, BerhardS. Online blind deconvolution for astronomical imaging [C]. IEEE International Conference on Computational Photography. San Francisco, USA, 20091-7

[2] DaubechiesI, DefriseM, MolC D. An iterative thresholding algorithm for linear inverse problems with a sparsity constraint [J]. Communications on Pure and Applied Mathematics, 2004, 57(11): 1413-1457

[3] MallatS, ZhangZ. Matching pursuit with time-frequency dictionary [J]. IEEE Transactions on Signal Processing, 1993, 41(12): 3397-3415

[4] CandesE, DonohoD L. Recovering edges in ill-posed inverse problems: Optimality of curvelet frames [J]. Analytical Statistics, 2002, 30(3): 784-842

[5] SunL, LuoJ-shu. Junk band recovery for hyperspectral image based on curvelet transform [J]. Journal of Central South University of Technology, 2011, 18: 816-822

[6] CandesE, DonohoD L. New tight frames of curvelets and the problem of approximating piecewise C2 images with piecewise C2 edges [J]. Communications on Pure and Applied Mathematics, 2004, 57(2): 219-266

[7] MairalJ, EladM, SapiroG. Sparse representation for color image restoration [J]. IEEE Transactions on Image Processing, 2008, 17(1): 53-69

[8] AharonM, EladM, BrucksteinA M. The K-SVD: An algorithm for designing of overcomplete dictionaries for sparse representation [J]. IEEE Transactions Signal Processing, 2006, 54(11): 4311-4322

[9] EladM, AharonM. Image denoising via sparse and redundant representations over learned dictionaries [J]. IEEE Transactions on Image Processing, 2006, 15(12): 489-509

[10] DongW, LiX, ZhangL, ShiG. Sparsity-based image denoising via dictionary learning and structural clustering [C]. IEEE Conference on Computer Vision and Pattern Recognition. Miami, USA, 2011457-464

[11] LiS, YinH, FangL. Group-sparse representation with dictionary learning for medical image denoising and fusion [J]. IEEE Transactions on Biomedical Engineering, 2012, 59(12): 3450-3459

[12] SmithL N, OlsonC C, JuddK P, NicholsJ M. Denoising infrared maritime imagery using tailored dictionaries via modified K-SVD algorithm [J]. Applied Optics, 2012, 51(17): 3941-3949

[13] RubinsteinR, ZibulevskyM, EladM. Double sparsity: Learning sparse dictionaries for sparse signal approximation [J]. IEEE Transactions on Signal Processing, 2010, 58(3): 1553-1564

[14] ObradovicR, JanevM, AnticB, CrnojevicaV, PetrovicN I. Robust sparse image denoising [C]. IEEE International Conference on Image Processing. Brussels, Belgium, 20112569-2572

[15] WangC-guangThe ECG feature wave detection and data compression based on the sparse decomposition [D], 2009ChangshaNational University of Defense Technology

[16] ZibulevskyM, EladM. L1–L2 optimization in signal and image processing [J]. IEEE Signal Processing Magazine, 2010, 27(3): 76-88

[17] DonohoD L. For most large underdetermined systems of linear equations the minimal L1-norm solution is also the sparsest solution [J]. Communications on Pure and Applied Mathematics, 2006, 59(6): 797-829

[18] LeeH, BattleA, RainaR, NgA. Efficient sparse coding algorithms [C]. Advances in Neural Information Processing Systems. Vancouver, Canada, 2007801-808

[19] YuK, ZhangT, GongY. Nonlinear learning using local coordinate coding [C]. Advances in Neural Information Processing Systems. Vancouver, Canada, 20092223-2231

[20] WangJ, YangJ, YuK, LvF, HuangT, GongY. Locality-constrained linear coding for image classification [C]. IEEE Conference on Computer Vision and Pattern Recognition. San Francisco, USA, 20103360-3367

[21] BeckA, TeboulleM. A fast iterative shrinkage thresholding algorithm for linear inverse problems [J]. SIAM Journal on Imaging Sciences, 2009183-202

[22] DuChunResearch on image motion deblurring and wavelet threshold denoising algorithm [D], 2008ChangshaNational University of Defense Technology