Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
194 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
46 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Noise2Noise Denoising of CRISM Hyperspectral Data (2403.17757v1)

Published 26 Mar 2024 in cs.CV and cs.LG

Abstract: Hyperspectral data acquired by the Compact Reconnaissance Imaging Spectrometer for Mars (CRISM) have allowed for unparalleled mapping of the surface mineralogy of Mars. Due to sensor degradation over time, a significant portion of the recently acquired data is considered unusable. Here a new data-driven model architecture, Noise2Noise4Mars (N2N4M), is introduced to remove noise from CRISM images. Our model is self-supervised and does not require zero-noise target data, making it well suited for use in Planetary Science applications where high quality labelled data is scarce. We demonstrate its strong performance on synthetic-noise data and CRISM images, and its impact on downstream classification performance, outperforming benchmark methods on most metrics. This allows for detailed analysis for critical sites of interest on the Martian surface, including proposed lander sites.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (25)
  1. Description of CoTCAT (Complement to CRISM Analysis Toolkit). IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 8(6):3039–3049, June 2015. ISSN 2151-1535. doi: 10.1109/JSTARS.2015.2405095.
  2. Automated processing of planetary hyperspectral datasets for the extraction of weak mineral signatures and applications to CRISM observations of hydrated silicates on Mars. Planetary and Space Science, 76:53–67, December 2012. ISSN 0032-0633. doi: 10.1016/j.pss.2012.11.007. URL https://www.sciencedirect.com/science/article/pii/S0032063312003625.
  3. Hydrous minerals on Mars as seen by the CRISM and OMEGA imaging spectrometers: Updated global view. Journal of Geophysical Research: Planets, 118(4):831–858, April 2013. ISSN 2169-9100. doi: 10.1029/2012JE004145. URL https://onlinelibrary.wiley.com/doi/abs/10.1029/2012JE004145. _eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1029/2012JE004145.
  4. Generation and Optimization of Spectral Cluster Maps to Enable Data Fusion of CaSSIS and CRISM Datasets. Remote Sensing, 14(11):2524, January 2022. ISSN 2072-4292. doi: 10.3390/rs14112524. URL https://www.mdpi.com/2072-4292/14/11/2524. Number: 11 Publisher: Multidisciplinary Digital Publishing Institute.
  5. Lovedeep Gondara. Medical Image Denoising Using Convolutional Denoising Autoencoders. In 2016 IEEE 16th International Conference on Data Mining Workshops (ICDMW), pp.  241–246, December 2016. doi: 10.1109/ICDMW.2016.0041. URL https://ieeexplore.ieee.org/abstract/document/7836672. ISSN: 2375-9259.
  6. Peter A. Gorry. General least-squares smoothing and differentiation by the convolution (Savitzky-Golay) method. Analytical Chemistry, 62(6):570–573, March 1990. ISSN 0003-2700. doi: 10.1021/ac00205a007. URL https://doi.org/10.1021/ac00205a007. Publisher: American Chemical Society.
  7. Quantitative Reconstruction and Denoising Method HyBER for Hyperspectral Image Data and Its Application to CRISM. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 12(4):1219–1230, April 2019. ISSN 2151-1535. doi: 10.1109/JSTARS.2019.2900644. URL https://ieeexplore.ieee.org/document/8672185.
  8. A new method for atmospheric correction and de-noising of CRISM hyperspectral data. Icarus, 354:114024, July 2020. ISSN 0019-1035. doi: 10.1016/j.icarus.2020.114024. URL https://www.sciencedirect.com/science/article/pii/S0019103520303857.
  9. Denoising Diffusion Restoration Models. Advances in Neural Information Processing Systems, 35:23593–23606, December 2022. URL https://proceedings.neurips.cc/paper_files/paper/2022/hash/95504595b6169131b6ed6cd72eb05616-Abstract-Conference.html.
  10. Regularization of Mars Reconnaissance Orbiter CRISM along‐track oversampled hyperspectral imaging observations of Mars. Icarus, 282:136–151, January 2017. ISSN 0019-1035. doi: 10.1016/j.icarus.2016.09.033. URL https://www.sciencedirect.com/science/article/pii/S0019103516301579.
  11. Noise2Noise: Learning Image Restoration without Clean Data, October 2018. URL http://arxiv.org/abs/1803.04189. arXiv:1803.04189 [cs, stat].
  12. Morphological and Spectral Diversity of the Clay-Bearing Unit at the ExoMars Landing Site Oxia Planum. Astrobiology, 21(4):464–480, April 2021. ISSN 1531-1074. doi: 10.1089/ast.2020.2292. URL https://www.liebertpub.com/doi/10.1089/ast.2020.2292. Publisher: Mary Ann Liebert, Inc., publishers.
  13. MRO/CRISM Retrieval of Surface Lambert Albedos for Multispectral Mapping of Mars With DISORT-Based Radiative Transfer Modeling: Phase 1—Using Historical Climatology for Temperatures, Aerosol Optical Depths, and Atmospheric Pressures. IEEE Transactions on Geoscience and Remote Sensing, 46(12):4020–4040, December 2008. ISSN 0196-2892. doi: 10.1109/TGRS.2008.2000631. URL http://ieeexplore.ieee.org/document/4685531/.
  14. CRISM Data Product Software Interface Specification Version 1.3.7.6., June 2022.
  15. Compact Reconnaissance Imaging Spectrometer for Mars investigation and data set from the Mars Reconnaissance Orbiter’s primary science phase. Journal of Geophysical Research: Planets, 114(E2), 2009. ISSN 2156-2202. doi: 10.1029/2009JE003344. URL https://onlinelibrary.wiley.com/doi/abs/10.1029/2009JE003344. _eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1029/2009JE003344.
  16. Visible to Short-Wave Infrared Spectral Analyses of Mars from Orbit Using CRISM and OMEGA. In Janice L. Bishop, James F. Bell III, and Jeffrey E. Moersch (eds.), Remote Compositional Analysis, pp.  453–483. Cambridge University Press, 1 edition, November 2019. ISBN 978-1-316-88887-2 978-1-107-18620-0. doi: 10.1017/9781316888872.025. URL https://www.cambridge.org/core/product/identifier/9781316888872%23CN-bp-23/type/book_part.
  17. Mario Parente. A New Approach To Denoise CRISM Images. In 39th Lunar and Planetary Science Conference, March 2008. URL https://www.lpi.usra.edu/meetings/lpsc2008/pdf/2528.pdf.
  18. A machine learning toolkit for CRISM image analysis. Icarus, 376:114849, April 2022. ISSN 00191035. doi: 10.1016/j.icarus.2021.114849. URL https://linkinghub.elsevier.com/retrieve/pii/S0019103521004905.
  19. Spectral Clustering of CRISM Datasets in Jezero Crater Using UMAP and k-Means. Remote Sensing, 15(4):939, January 2023. ISSN 2072-4292. doi: 10.3390/rs15040939. URL https://www.mdpi.com/2072-4292/15/4/939. Number: 4 Publisher: Multidisciplinary Digital Publishing Institute.
  20. MarsSI: Martian surface data processing information system. Planetary and Space Science, 150:157–170, January 2018. ISSN 0032-0633. doi: 10.1016/j.pss.2017.09.014. URL https://www.sciencedirect.com/science/article/pii/S0032063316304718.
  21. Data Processing and Analysis Products Update - Calibration, Correction, and Visualisation. In 42nd Lunar and Planetary Sciences Conference, 2011.
  22. The CRISM investigation in Mars orbit: Overview, history, and delivered data products. Icarus, pp.  115612, May 2023. ISSN 0019-1035. doi: 10.1016/j.icarus.2023.115612. URL https://www.sciencedirect.com/science/article/pii/S0019103523001896.
  23. Deep learning on image denoising: An overview. Neural Networks, 131:251–275, November 2020. ISSN 0893-6080. doi: 10.1016/j.neunet.2020.07.025. URL https://www.sciencedirect.com/science/article/pii/S0893608020302665.
  24. Revised CRISM spectral parameters and summary products based on the currently detected mineral diversity on Mars. Journal of Geophysical Research: Planets, 119(6):1403–1431, 2014. ISSN 2169-9100. doi: 10.1002/2014JE004627. URL https://onlinelibrary.wiley.com/doi/abs/10.1002/2014JE004627. _eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/2014JE004627.
  25. FFDNet: Toward a Fast and Flexible Solution for CNN-Based Image Denoising. IEEE Transactions on Image Processing, 27(9):4608–4622, September 2018. ISSN 1941-0042. doi: 10.1109/TIP.2018.2839891. URL https://ieeexplore.ieee.org/abstract/document/8365806. Conference Name: IEEE Transactions on Image Processing.

Summary

We haven't generated a summary for this paper yet.

X Twitter Logo Streamline Icon: https://streamlinehq.com