Abstract
In this paper, we introduce an efficient algorithm for reconstructing incomplete images based on optimal least-squares (LS) approximation. Generally, LS method requires a low-rank basis set that can represent the overall characteristic of an image, which can be obtained optimally via the singular value decomposition (SVD). This basis is called proper orthogonal decomposition (POD) basis. To significantly decrease the computational cost of SVD, this work employs a randomized singular value decomposition (rSVD) to compute the basis from the available image pixels. In this work, to preserve the 2-dimensional structure of the image, the test image is first subdivided into many 2-dimensional small patches. The complete patches are used to compute the POD basis for reconstructing corrupted patches. For each incomplete patch, the known pixels in the neighborhood around the missing components are used in the LS approximation together with the POD basis in the reconstruction process. The numerical tests compare the execution time used in computing this optimal low-rank basis by using rSVD and SVD, as well as demonstrate the accuracy of the resulting image reconstructions.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Emmanuel, J.C., Benjamin, R.: Exact matrix completion via convex optimization. Found. Comput. Math. 9, 717–772 (2008)
Jin, Z., Chiman, K., Bulent, A.: A high performance missing pixel reconstruction algorithm for hyperspectral images. In: Proceedings of 2nd International Conference on Applied and Theoretical Information Systems Research (2012)
Feilong, C., Miaaomiao, C., Yuanpeng, T.: Image interpolation via low-rank matrix completion and recovery. IEEE Trans. Circuits Syst. Video Technol. 25(8), 1261–1270 (2015)
Pete, G.W.S.: On the early history of the singular value decomposition. Siam Rev. 35, 551–566 (1993) http://www.jstor.org/stable/2132388
Dan, K.: A singularly valuable decomposition: the svd of a matrix. The College Math. J. 27(1), 2–23 (1996)
Dimitris, A., Frank, M.: Fast computation of low-rank matrix approximations, J. Assoc. Comput. Mach., 54 (2007)
Gilbert, S.: Linear Algebra and its Applications, 3rd edn. Harcourt Brace Jovanovich, San Diego, USA (1988)
Gene, H.G., Charies, F.V.L.: Matrix Computations, 3rd ed., Johns Hopkins University Press, Baltimore, Maryland (1996)
Harry, C.A., Claude, L.P.: Singular value decompositions and digital image processing, 24, 26–53 (1976)
Samruddh, K., Reena, R.: Image compression using singular value decomposition. Int. J. Adv. Res. Technol. 2(8), 244–248 (2013)
Lijie, C.: Singular Value Decomposition Applied To Digital Image Processing, Division of Computing Studies. Arizona State University Polytechnic Campus, Mesa, Arizona, 1–15
Sven, O.A., John, H.H., Patrick, W.: A critique of SVD-based image coding systems. In: IEEE International Symposium on Circuits and Systems (ISCAS), vol. 4, pp. 13–16 (1999)
Vitali, V.S., Roger, L.: Fast PET image reconstruction based on svd decomposition of the system matrix. IEEE Trans. Nuclear Sci. 48(3), 761–767 (2001)
Rowayda, A.S.: SVD based image processing applications: state of the art, contributions and research challenges. Int. J. Adv. Comput. Sci. and Appl. 3(7), 26–34 (2012)
Davi, M.L., João, P.C.L.C., João, L.A.C.: Improved MRI reconstruction and denoising using SVD-based low-rank approximation. In: Workshop on Engineering Applications, pp. 1–6 (2012)
Sarlos, T.: Improved approximation algorithms for large matrices via random projections. In: 47th Annual IEEE Symposium Foundations of Computer Science, pp. 143–152(2006)
Liberty, E., Woolfe, F., Martinsson, P.G., Rokhlin, V., Tygert, M.: Randomized algorithms for the low-rank approximation of matrices. Proc. Natl. Acad. Sci. 104(51), 20167–20172 (2007)
Halko, N., Martinsson, P.G., Tropp, J.A.: Finding structure with randomness: probabilistic algorithms for constructing approximate matrix decompositions. Siam Rev. 53(2), 217–288 (2011)
Siriwan, I., Saifon, C.: An application of randomized singular value decomposition on image reconstruction using least-squares approach, Preprint in Thai Journal of Mathematics (2018)
Siriwan, I., Saifon, C.: A Numerical Study of Efficient Sampling Strategies for Randomized Singular Value Decomposition, Thai Journal of Mathematics, pp. 351–370 (2019)
Acknowledgments
The authors gratefully acknowledge the financial support provided by this study was supported by Thammasat University Research Fund, Contract No. TUGR 2/12/2562 and Royal Thai Government Scholarship in the Area of Science and Technology (Ministry of Science and Technology).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Intawichai, S., Chaturantabut, S. (2021). Missing Image Data Reconstruction Based on Least-Squares Approach with Randomized SVD. In: Vasant, P., Zelinka, I., Weber, GW. (eds) Intelligent Computing and Optimization. ICO 2020. Advances in Intelligent Systems and Computing, vol 1324. Springer, Cham. https://doi.org/10.1007/978-3-030-68154-8_89
Download citation
DOI: https://doi.org/10.1007/978-3-030-68154-8_89
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-68153-1
Online ISBN: 978-3-030-68154-8
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)