Abstract
Hyperspectral images (HSIs) frequently experience various types of noise due to atmospheric interference and sensor instability, which impairs the efficiency of subsequent operations. Consequently, HSI denoising has emerged as a crucial component of HSI preprocessing. Conventional approaches often target a single kind of noise and eliminate it repeatedly, which has disadvantages including inefficiency when handling heterogeneous noise. Lately, models based on deep neural networks have shown encouraging results in the general image denoising domain. This study, which aims to overcome shortcomings in previous techniques, provides a novel denoising methodology by leveraging the effectiveness of the SqueezeNet model. For a thorough assessment, the evaluation framework includes four main indicators: PSNR, SSIM, SAM, and ERGAS. The evaluation is based on real-world hyperspectral images from the [Harvard Hyperspectral Dataset], which cover a variety of scenarios and illumination circumstances. Fire blocks are used by the SqueezeNet-based denoising model to optimize feature extraction with fewer parameters.Benchmarks for comparison include deep learning technique QRNN3D and classical techniques like ITSReg and BM4D.In order to avoid convergence to suboptimal local minima and to speed up and stabilize the learning process, this work presents an incremental training policy. The suggested SqueezeNet-based HSI denoising model performs exceptionally well, attaining competitive results in terms of PSNR of 34.15, SSIM of 0.92, and SAM of 4.56 in addition to impressive ERGAS of 20.47. This study offers an effective denoising solution for hyperspectral images by addressing shortcomings in current techniques, showcasing improvements in efficiency and accuracy.
Data availability
The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.
References
Aslam, M. A., Ali, M. T., Nawaz, S., Shahzadi, S. & Fazal, M. A. Classification of Rethinking Hyperspectral Images Using 2D and 3D CNN with Channel and Spatial Attention: A Review.
Kucuk, S. & Yuksel, S. E. Total utility metric-based dictionary pruning forsparse hyperspectral unmixing. IEEE Trans. Comput. Imaging. 7, 562–572 (2021).
Chaturvedi, R. et al. Ultrafast Hot Carrier Injection in core–shell Au@MoS₂ Systems for plasmonic-assisted Hydrogen Evolution 1–13 (Plasmonics, 2025). https://doi.org/10.1007/s11468-025-03190-2
Sharma, A. et al. Ultrafast Carrier Dynamics in plasmonic-enhanced Perovskite Thin Films for next-gen solar-powered Wearables 1–11 (Plasmonics, 2025). https://doi.org/10.1007/s11468-025-03221-y
Singh, R. P. et al. AI-driven Discovery of Alloyed Plasmonic Nanodisks for Broadband Solar Absorption and Charge Carrier Multiplication 1–14 (Plasmonics, 2025). https://doi.org/10.1007/s11468-025-03223-w
Singh, R. P. et al. Self-guided Assembly of ligand-free Rh–Os Nanotrees Revealing Plasmonic Branching pathways, tip-enhanced Field localization, and femtosecond-induced third-order Nonlinear Optical Effects 1–14 (Plasmonics, 2025). https://doi.org/10.1007/s11468-025-03259-y
Updhay, V. V. et al. Temporal Coding of Incident Light on phase-change Plasmonic Surfaces for Adaptive Optical Memory Storage 1–14 (Plasmonics, 2025). https://doi.org/10.1007/s11468-025-03218-7
Updhay, V. V. et al. Graphene–plasmon hybrid interlayers for dynamically tunable hot electron generation in visible-to-NIR ranges. Plasmonics https://doi.org/10.1007/s11468-025-03260-5 (2025).
Sharma, A. et al. Multiphysics-guided design and on-chip integration of Ru–Pt nanocomposites for enhanced thermoelectric–plasmonic energy harvesting. Plasmonics. (2025). https://doi.org/10.1007/s11468-025-03210-1
Zhang, Q. et al. Hybrid noise removal in hyperspectral imagery with a spatial–spectral gradient network,IEEE Trans. Geosci. Remote Sens. 57 (10), 7317–7329 (2019).
Bahraini, T., Ebrahimi-Moghadam, A., Khademi, M. & Yazdi, H. S. Bayesianframework selection for hyperspectral image denoising. Signal. Process. 201, 108712 (2022).
Chen Chen, W. et al. Spectral–spatial preprocessingusing mult hypothesis prediction for noise-robust hyperspectralimage classification. IEEE J. SelectedTopics Appl. Earth Observations Remote Sens. 7 (4), 1047–1059 (2014).
Chen, H., Yang, G. & Zhang, H. Hider: A Hyperspectral Image Denoising Transformer with spatial–spectral Constraints for Hybrid Noise Removal (IEEE Transactions on Neural Networks and Learning Systems, 2022).
Vidal, M. & Amigo, J. Pre-processing of hyperspectral images.Essential steps before image analysis, Chemometrics Intell. Lab. Syst.,vol. 117, pp. 138–148, Aug. (2012).
Subudhi, S., Patro, R. N., Biswal, P. K. & Dell’Acqua, F. A survey on superpixel segmentation as a preprocessing step in hyperspectral image analysis. IEEE J. Sel. Top. Appl. Earth Observations Remote Sens. 14, 5015–5035 (2021).
Gu, S., Zhang, L., Zuo, W. & Feng, X. Weighted nuclear norm minimization with application to image denoising. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 2862-2869). (2014).
Dabov, K., Foi, A., Katkovnik, V. & Egiazarian, K. Image denoisingby sparse 3-D transform-domain collaborative filtering, IEEE Trans.Image Process., vol. 16, no. 8, pp. 2080–2095, Aug. (2007). https://doi.org/10.1109/TIP.2007.901238
Li, M., Fu, Y. & Zhang, Y. Spatial-spectral transformer for hyperspectral image denoising. In Proceedings of the AAAI Conference on Artificial Intelligence (Vol. 37, No. 1, pp. 1368-1376). (2023), June.
De Oliveira, G. A., Almeida, L. M., De Lima, E. R. & Meloni, L. G. P. Deep Convolutional Network Aided by Non-Local Method for Hyperspectral Image Denoising (IEEE Access, 2023).
Gkillas, A., Ampeliotis, D. & Berberidis, K. Connections between deep equilibrium and sparse representation models with application to hyperspectral image denoising. IEEE Trans. Image Process. 32, 1513–1528 (2023).
Xiong, F. et al. SMDS-Net: model guided spectral-spatial network for hyperspectral image denoising. IEEE Trans. Image Process. 31, 5469–5483 (2022).
Pang, L., Gu, W. & Cao, X. TRQ3DNet: A 3D quasi-recurrent and transformer-based network for hyperspectral image denoising. Remote Sens. 14 (18), 4598 (2022).
Aetesam, H., Maji, S. K. & Yahia, H. Bayesian approach in a learning-based hyperspectral image denoising framework. IEEE Access. 9, 169335–169347 (2021).
Updhay, V. V. et al. Single-cell photothermal ablation using wavelength-matched Hollow gold nanostars with adaptive feedback control. Plasmonics https://doi.org/10.1007/s11468-025-03217-8 (2025).
Updhay, V. V. et al. Neuromorphic integration and real-time programmability of temporally-coded phase-change plasmonic platforms for on-chip multilevel optical memory and adaptive logic systems. Micro Nanostruct. 208, 208317. https://doi.org/10.1016/j.micrna.2025.208317 (2025).
Iandola, F. N. et al. SqueezeNet: AlexNet-level accuracy with 50x fewerparameters and <0.5 MB model size. arXiv arXiv:1602.07360. (2016).
Agarap, A. F. Deep learning using rectified linear units (relu). arXiv preprint arXiv:1803.08375. (2018).
Ahissarand, M. & Hochstein, S. The reverse hierarchy theory of visualperceptual learning. Trends Cogn. Sci. 8 (10), 457–464 (2004).
Wang, Z., Bovik, A. C. & Sheikh, H. R. Simoncelli. Imagequality assessment: from error visibility to structural similarity. IEEETransactions Image Process. 13 (4), 600–612 (2004).
Yuhas, R. H., Boardman, J. W. & Goetz, A. F. Determination of semiaridlandscape endmembers and seasonal trends using convex geometryspectral unmixing techniques. In Summaries of the 4-th Annual JPLAirborne Geoscience Workshop, (1993).
Wald, L. Data Fusion: Definitions Architectures: Fusion Images DifferentSpatial Resolutions (Presses des MINES, 2002).
Ayan, C. & Zickler, T. Statistics of Real-World Hyperspectral Images, in Proceedingsof the IEEE Conference on Computer Vision and Pattern Recognition (CVPR),(2011).
Ranjan, P., Kumar, R. & Girdhar, A. A 3D-convolutional-autoencoder embedded Siamese-attention-network for classification of hyperspectral images. Neural Comput. Appl. 36, 8335–8354. https://doi.org/10.1007/s00521-024-09527-y (2024).
Ranjan, P., Kaushal, A., Girdhar, A. & Kumar, R. Revolutionizing hyperspectral image classification for limited labeled data: unifying autoencoder-enhanced GANs with convolutional neural networks and zero-shot learning. Earth Sci. Inf. (2). https://doi.org/10.1007/s12145-025-01739-7 (2025). Springer Science and Business Media LLC.
Ranjan, P., Kumar, R. & Girdhar, A. Recent CNN advancements for stratification of hyperspectral images. 6th International Conference on Information Systems and Computer Networks (ISCON), Mathura, India, 2023, pp. 1–5., Mathura, India, 2023, pp. 1–5. (2023). https://doi.org/10.1109/ISCON57294.2023.10112174
Ranjan, P. & Girdhar, A. A comparison of deep learning algorithms dealing with limited samples in hyperspectral image classification. OPJU International Technology Conference on Emerging Technologies for Sustainable Development (OTCON), Raigarh, Chhattisgarh, India, 2023, pp. 1–6., Raigarh, Chhattisgarh, India, 2023, pp. 1–6. (2022). https://doi.org/10.1109/OTCON56053.2023.10114005
Ranjan, P., Kumar, R. & Girdhar, A. Unlocking the potential of unlabeled data: semi-supervised learning for stratification of hyperspectral images. OITS International Conference on Information Technology (OCIT), Raipur, India, 2023, pp. 938–943., Raipur, India, 2023, pp. 938–943. (2023). https://doi.org/10.1109/OCIT59427.2023.10430513
Ranjan, P., Kumar, R. & Jung, K. H. Exploring cutting edge of AI in hyperspectral image classification. Proceedings of the 2024 Summer Conference of the Korean Institute of Electrical Engineers (KIEE), pp. 2525–2530. (2024).
Ranjan, P. & Girdhar, A. Deep Siamese network with handcrafted feature extraction for hyperspectral image classification. Multimedia Tools Appl. 83, 2501–2526. https://doi.org/10.1007/s11042-023-15444-4 (2024).
Ranjan, P. & Girdhar, A. A comprehensive systematic review of deep learning methods for hyperspectral image classification. Int. J. Remote Sens. 43 (17), 6221–6306. https://doi.org/10.1080/01431161.2022.2133579 (2022).
Ranjan, P. et al. A novel spectral-spatial 3D auxiliary conditional GAN integrated convolutional LSTM for hyperspectral image classification. Earth Sci. Inf. 17, 5251–5271. https://doi.org/10.1007/s12145-024-01451-y (2024).
Ranjan, P. & Gupta, G. A cross-domain semi-supervised zero-shot learning model for the classification of hyperspectral images. J. Indian Soc. Remote Sens. 51, 1991–2005. https://doi.org/10.1007/s12524-023-01734-9 (2023).
Ranjan, P. & Girdhar, A. Xcep-Dense: a novel lightweight extreme inception model for hyperspectral image classification. Int. J. Remote Sens. 43 (14), 5204–5230. https://doi.org/10.1080/01431161.2022.2130727 (2022).
Acknowledgements
There is no acknowledgement involved in this work.
Funding
No funding is involved in this work.
Author information
Authors and Affiliations
Contributions
Dr. Nandhagopal Nachimuthu - Writing- original draft, ConceptualizationRAMYA MURUGESAN - Writing- review & editing, Data CurationDr.M.Dharmalingam - Formal analysis, Funding acquisitionDr.G.PRAKASH – Investigation, Methodology.
Corresponding author
Ethics declarations
Ethics approval and consent to participate
No participation of humans takes place in this implementation process.
Competing interests
The authors declare no competing interests.
Human and animal rights
No violation of Human and Animal Rights is involved.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Nachimuthu, N., Murugesan, R., Dharmalingam, M. et al. Revolutionizing hyper spectral image denoising: a squeezenet paradigm. Sci Rep (2026). https://doi.org/10.1038/s41598-026-36479-1
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-026-36479-1
