Abstract
Many applications on hyperspectral images (HSIs) always suffer from the high dimension low sample size problem. Hence dimension reduction (DR) is a necessary pre-process for HSIs. Most existing DR algorithms only concentrate on a single dataset. However, similar HSI scenes may share information between each other. How to utilize the shared information is an interesting research topic. This research work concentrates on cross-scene DR for HSIs. Combining the idea of dual dictionary non-negative matrix factorization (DDNMF) and stacked autoencoder (SAE), a multi-layer cross-domain non-negative matrix factorization (MLCDNMF) algorithm is proposed in this work to perform DR across different HSI scenes. MLCDNMF has following characters: 1) it provides the ability of both homogenous transfer learning and heterogeneous transfer learning; 2) it containers two DDNMF layers, and works in a way like SAE; 3) it is beyond SAE due to the graphs built based on the correlations between samples; 4) it is a flexible model which does not require strict one-to-one sample correspondences between different scenes. Experiments on cross-scene HSI datasets show good performance of the proposed MLCDNMF algorithm.
Supported by the National Natural Science Foundation of China (grant number 61701468), the National Key Research and Development Program of China (grant number 2018YFB0505000) and the Outstanding Student Achievement Cultivation Program of China Jiliang University (grant number 2019YW24).
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsReferences
Agarwal, A., El-Ghazawi, T., El-Askary, H., Le-Moigne, J.: Efficient hierarchical-PCA dimension reduction for hyperspectral imagery. In: Proceedings pf the IEEE International Symposium on Signal Processing and Information Technology, pp. 353–356, December 2007
Chen, H., Ye, M., Lu, H., Lei, L., Qian, Y.: Dual dictionary learning for mining a unified feature subspace between different hyperspectral image scenes. In: Proceedings of the IEEE International Geoscience and Remote Sensing Symposium, pp. 1096–1099 (2019)
Hinton, G.E., Zemel, R.S.: Autoencoders, minimum description length and Helmholtz free energy. In: Proceedings of the Advances in Neural Information Processing Systems, pp. 3–10 (1994)
Li, W., Feng, F., Li, H., Du, Q.: Discriminant analysis-based dimension reduction for hyperspectral image classification: a survey of the most recent advances and an experimental comparison of different techniques. IEEE Geosci. Remote Sens. Mag. 6(1), 15–34 (2018)
Wang, C., Mahadevan, S.: A general framework for manifold alignment. In: Proceedings of the AAAI Fall Symposium Series, pp. 79–86 (2009)
Ye, M., Qian, Y., Zhou, J., Tang, Y.Y.: Dictionary learning-based feature-level domain adaptation for cross-scene hyperspectral image classification. IEEE Trans. Geosci. Remote Sens. 55(3), 1544–1562 (2017)
Ye, M., Zheng, W., Lu, H., Zeng, X., Qian, Y.: Cross-scene hyperspectral image classification based on DWT and manifold-constrained subspace learning. Int. J. Wavelets Multiresolution Inf. Process. 15(06), 1–16 (2017)
Zabalza, J., Ren, J., Zheng, J., Zhao, H., Qing, C., Yang, Z., Du, P., Marshall, S.: Novel segmented stacked autoencoder for effective dimensionality reduction and feature extraction in hyperspectral imaging. Neurocomputing 185, 1–10 (2016)
Zhao, W., Du, S.: Spectral-spatial feature extraction for hyperspectral image classification: a dimension reduction and deep learning approach. IEEE Trans. Geosci. Remote Sens. 54(8), 4544–4554 (2016)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Switzerland AG
About this paper
Cite this paper
Chen, H., Gong, K., Lei, L., Ye, M., Qian, Y. (2020). Multi-layer Cross-domain Non-negative Matrix Factorization for Cross-scene Dimension Reduction on Hyperspectral Images. In: Lu, Y., Vincent, N., Yuen, P.C., Zheng, WS., Cheriet, F., Suen, C.Y. (eds) Pattern Recognition and Artificial Intelligence. ICPRAI 2020. Lecture Notes in Computer Science(), vol 12068. Springer, Cham. https://doi.org/10.1007/978-3-030-59830-3_47
Download citation
DOI: https://doi.org/10.1007/978-3-030-59830-3_47
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-59829-7
Online ISBN: 978-3-030-59830-3
eBook Packages: Computer ScienceComputer Science (R0)