Abstract
Low-rank and sparsity are two popular regularization objects in hyperspectral image restoration. To generalize this decomposition, a weighted non-convex low-rank method is proposed by adopting non-convex \(l_p\)-norm (\(0<p<1\)) to recover the clean image more genuinely. The noise term including Gaussian noise, impulse noise, stripes, and deadlines, is assumed to be sparse, and the non-convex \(l_p\)-norm can better approach the \(l_0\)-norm when \(0<p<1\). For the clean image underlying a noised one, its low rank property is conducive to the capture of both spatial and spectral information by applying \(l_p\)-norm to the bases of the gradient of data. To remain faithful to the dissimilarity between the correlation along spatial and spectral directions, weights are assigned when computing the difference in each direction. Although being non-convex, the proposed model can be solved via ADMM algorithm. Extensive simulated experiments show that this model outperforms various existing methods visually and quantitatively.
Keywords
This work was supported in part by the National Nature Science Foundation of China (62072312, 61972264, 61872429, 61772343), in part by Shenzhen Basis Research Project (JCYJ20180305125521534, JCYJ20170818091621856, JCYJ20170302144838601), in part by the Interdisciplinary Innovation Team of Shenzhen University (SZUGS2021SEMINAR06, JG2020060), in part by National Nature Science Foundation of Guangdong province (2019A1515010894).
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Casorati matrix is a matrix whose columns comprise vectorized bands of an HSI [5].
- 2.
References
Goetz, A.F.: Three decades of hyperspectral remote sensing of the earth: a personal view. Remote Sens. Environ. 113, S5–S16 (2009)
Goldfarb, D., Osher, S., Burger, M., Xu, J., Yin, W.: An iterative regularization method for total variation-based image restoration. Multiscale Model. Simul. 4(2), 460–489 (2005)
Gu, S., Zhang, L., Zuo, W., Feng, X.: Weighted nuclear norm minimization with application to image denoising. In: 2014 IEEE Conference on Computer Vision and Pattern Recognition, pp. 2862–2869 (2014)
He, W., Zhang, H., Zhang, L., Shen, H.: Hyperspectral image denoising via noise-adjusted iterative low-rank matrix approximation. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 8(6), 3050–3061 (2015)
He, W., Zhang, H., Zhang, L., Shen, H.: Total-variation-regularized low-rank matrix factorization for hyperspectral image restoration. IEEE Trans. Geosci. Remote Sens. 54(1), 178–188 (2016)
Krishnan, D., Fergus, R.: Fast image deconvolution using hyper-laplacian priors. In: Bengio, Y., Schuurmans, D., Lafferty, J.D., Williams, C.K.I., Culotta, A. (eds.) Advances in Neural Information Processing Systems 22: 23rd Annual Conference on Neural Information Processing Systems 2009. Proceedings of a Meeting Held 7–10 December 2009, Vancouver, British Columbia, Canada, pp. 1033–1041. Curran Associates, Inc. (2009)
Lu, C., Feng, J., Chen, Y., Liu, W., Lin, Z., Yan, S.: Tensor robust principal component analysis with a new tensor nuclear norm. IEEE Trans. Pattern Anal. Mach. Intell. 42(4), 925–938 (2020)
Peng, J., Xie, Q., Zhao, Q., Wang, Y., Yee, L., Meng, D.: Enhanced 3DTV regularization and its applications on HSI denoising and compressed sensing. IEEE Trans. Image Process. 29, 7889–7903 (2020)
Rudin, L.I., Osher, S., Fatemi, E.: Nonlinear total variation based noise removal algorithms. Physica D 60(1), 259–268 (1992)
Wang, Y., Peng, J., Zhao, Q., Leung, Y., Zhao, X., Meng, D.: Hyperspectral image restoration via total variation regularized low-rank tensor decomposition. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 11(4), 1227–1243 (2018)
Wright, J., Ganesh, A., Rao, S.R., Peng, Y., Ma, Y.: Robust principal component analysis: exact recovery of corrupted low-rank matrices via convex optimization. In: Bengio, Y., Schuurmans, D., Lafferty, J.D., Williams, C.K.I., Culotta, A. (eds.) Advances in Neural Information Processing Systems 22: 23rd Annual Conference on Neural Information Processing Systems 2009. Proceedings of a Meeting Held 7–10 December 2009, Vancouver, British Columbia, Canada, pp. 2080–2088. Curran Associates, Inc. (2009)
Xie, Q., et al.: Multispectral images denoising by intrinsic tensor sparsity regularization. In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 1692–1700 (2016)
Xie, Y., Qu, Y., Tao, D., Wu, W., Yuan, Q., Zhang, W.: Hyperspectral image restoration via iteratively regularized weighted Schatten \(p\)-norm minimization. IEEE Trans. Geosci. Remote Sens. 54(8), 4642–4659 (2016)
Zhang, H., He, W., Zhang, L., Shen, H., Yuan, Q.: Hyperspectral image restoration using low-rank matrix recovery. IEEE Trans. Geosci. Remote Sens. 52(8), 4729–4743 (2014)
Zhou, T., Tao, D.: GoDec: randomized low-rank and sparse matrix decomposition in noisy case. In: Proceeding of the Twenty-Eighth International Conference on Machine Learning, pp. 33–40. International Machine Learning Society (IMLS) (2011)
Fan, J., Li, R.: Variable selection via nonconcave penalized likelihood and its Oracle properties. J. Am. Stat. Assoc. 96, 1348–1361 (2001)
Nie, F., Wang, H., Cai, X., Huang, H., Ding, C.: Robust matrix completion via joint Schatten \(p\)-norm and \(l_p\)-norm minimization. IEEE International Conference on Data Mining, Brussels, Belgium, pp. 566–574 (2012)
Shang, F., Liu, Y., Cheng, J.: Scalable algorithms for tractable Schatten quasi-norm minimization. In: Computing Research Repository. arXiv:1606.01245 (2016)
Xu, C., Lin, Z., Zha, H.: A unified convex surrogate for the Schatten-p norm. In: The Thirty-First AAAI Conference on Artificial Intelligence, San Francisco, California, USA, pp. 926–932 (2017)
Shang, F., Cheng, J., Liu, Y., Luo, Z., Lin, Z.: Bilinear linear factor matrix norm minimization for robust PCA: algorithms and applications. IEEE Trans. Pattern Anal. Mach. Intell. 50, 2066–2080 (2018)
Zuo, W., Meng, D., Zhang, L., Feng, X., Zhang, D.: A generalized iterated shrinkage algorithm for non-convex sparse coding. In: IEEE International Conference on Computer Vision, pp. 217–224 (2013)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Yu, Y., Li, M. (2021). A Weighted Non-convex Restoration Model for Hyperspectral Image. In: Peng, Y., Hu, SM., Gabbouj, M., Zhou, K., Elad, M., Xu, K. (eds) Image and Graphics. ICIG 2021. Lecture Notes in Computer Science(), vol 12889. Springer, Cham. https://doi.org/10.1007/978-3-030-87358-5_14
Download citation
DOI: https://doi.org/10.1007/978-3-030-87358-5_14
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-87357-8
Online ISBN: 978-3-030-87358-5
eBook Packages: Computer ScienceComputer Science (R0)