Abstract
Aiming at handling the problem caused by the lack of prior spectral knowledge of anomalous pixels for hyperspectral anomaly detection (HAD). In this paper, we propose a background augmentation with transformer-based autoencoder for hyperspectral remote sensing image anomaly detection. The representative background pixels are selected based on sparse representation for obtaining typical background pixels as training samples of the transformer-based autoencoder. The selected typical background pixels can be used for training the transformer-based autoencoder to realize background pixel reconstruction. Thereafter, the pseudo background samples can be reconstructed from the transformer-based autoencoder, which is used to subtract the original image to obtain the residual image. Finally, Reed-Xiaoli (RX) is used to detect the anomalous pixels from residual image. Experiments results demonstrate that the proposed transformer-based autoencoder which can present competitive hyperspectral image anomaly detection results than other traditional algorithms.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Su, H., Wu, Z., Zhang, H., et al.: Hyperspectral anomaly detection: a survey. IEEE Geosci. Remote Sens. Mag. 10(1), 64–90 (2021)
Reed, I.S., Yu, X.: Adaptive multiple-band CFAR detection of an optical pattern with unknown spectral distribution. IEEE Trans. Acoust. Speech Signal Process. 38(10), 1760–1770 (1990)
Matteoli, S., Veracini, T., Diani, M., et al.: A locally adaptive background density estimator: an evolution for RX-based anomaly detectors. IEEE Geosci. Remote Sens. Lett. 11(1), 323–327 (2013)
Li, W., Du, Q.: Collaborative representation for hyperspectral anomaly detection. IEEE Trans. Geosci. Remote Sens. 53(3), 1463–1474 (2014)
Li, L., Li, W., Qu, Y., et al.: Prior-based tensor approximation for anomaly detection in hyperspectral imagery. IEEE Trans. Neural Netw. Learn. Syst. (2020)
Fu, X., Jia, S., Zhuang, L., et al.: Hyperspectral anomaly detection via deep plug-and-play denoising CNN regularization. IEEE Trans. Geosci. Remote Sens. 59(11), 9553–9568 (2021)
Ma, N., Peng, Y., Wang, S., et al.: An unsupervised deep hyperspectral anomaly detector. Sensors 18(3), 693 (2018)
Lu, X., Zhang, W., Huang, J.: Exploiting embedding manifold of autoencoders for hyperspectral anomaly detection. IEEE Trans. Geosci. Remote Sens. 58(3), 1527–1537 (2019)
Jiang, T., Li, Y., Xie, W., et al.: Discriminative reconstruction constrained generative adversarial network for hyperspectral anomaly detection. IEEE Trans. Geosci. Remote Sens. 58(7), 4666–4679 (2020)
Vaswani, A., Shazeer, N., Parmar, N., et al.: Attention is all you need. In: Advances in Neural Information Processing Systems, vol. 30 (2017)
Dosovitskiy, A., Beyer, L., Kolesnikov, A., et al.: An image is worth 16x16 words: transformers for image recognition at scale. arXiv preprint arXiv:2010.11929 (2020)
Fang, Y., Liao, B., Wang, X., et al.: You only look at one sequence: rethinking transformer in vision through object detection. In: Advances in Neural Information Processing Systems, vol. 34, pp. 26183–26197 (2021)
Abdi, H., Williams, L.J.: Principal component analysis. Wiley Interdiscip. Rev. Comput. Stat. 2(4), 433–459 (2010)
Tropp, J.A., Gilbert, A.C.: Signal recovery from random measurements via orthogonal matching pursuit. IEEE Trans. Inf. Theory 53(12), 4655–4666 (2007)
Acknowledgements
Manuscript received May 10, 2022, accepted June 1, 2022. This work was supported in part by the National Natural Science Foundation of China (No. 61801353), in part by The Project Supported by the China Postdoctoral Science Foundation funded project (No. 2018M633474), in part by GHfund under grant number 202107020822 and 202202022633.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 IFIP International Federation for Information Processing
About this paper
Cite this paper
Wang, J., Liu, Y., Li, L. (2022). Background Augmentation with Transformer-Based Autoencoder for Hyperspectral Anomaly Detection. In: Shi, Z., Jin, Y., Zhang, X. (eds) Intelligence Science IV. ICIS 2022. IFIP Advances in Information and Communication Technology, vol 659. Springer, Cham. https://doi.org/10.1007/978-3-031-14903-0_32
Download citation
DOI: https://doi.org/10.1007/978-3-031-14903-0_32
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-14902-3
Online ISBN: 978-3-031-14903-0
eBook Packages: Computer ScienceComputer Science (R0)
