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.

What problem does this paper attempt to address?

The paper primarily addresses the noise issue in hyperspectral imaging data of the Martian surface and proposes a new self-supervised deep learning model architecture—Noise2Noise4Mars (N2N4M). Specifically, the paper attempts to solve the following key problems: 1. **Decline in Data Quality**: With the increasing service time of the Mars Reconnaissance Orbiter (MRO), the sensor performance of the Compact Reconnaissance Imaging Spectrometer for Mars (CRISM) has gradually degraded, leading to a significant increase in noise levels in the data acquired in recent years, which affects the identification and analysis of mineral features. 2. **Noise Impact on Scientific Analysis**: The presence of noise makes detailed analysis of certain key locations (such as the Oxia Planum region on Mars, one of the candidate landing sites for the European Space Agency's ExoMars rover) difficult. 3. **Lack of High-Quality Labeled Data**: In the field of planetary science, especially for Martian hyperspectral data, it is challenging to obtain zero-noise target data as training samples, which limits the application of existing data-driven methods. To address the above issues, the researchers developed the N2N4M model, a self-supervised deep learning architecture for denoising CRISM images. This model does not require paired noiseless target data for training, making it highly suitable for processing hyperspectral data, particularly in planetary science applications. Moreover, the N2N4M model can handle target data that inherently contains noise, which is crucial for practical application scenarios where perfect data is rare. Experimental results show that the N2N4M model exhibits good denoising performance on both synthetic noise data and real CRISM images, demonstrating significant advantages over traditional methods (such as the Complement to CRISM Analysis Toolkit (CoTCAT)). Additionally, the effectiveness of the N2N4M model was further validated by applying the denoised data to downstream classification tasks. Overall, this study provides strong support for enhancing the capability of mineralogical analysis on the Martian surface.

Noise2Noise Denoising of CRISM Hyperspectral Data

Deep Learning-Based Super-Resolution Reconstruction and Algorithm Acceleration of Mars Hyperspectral CRISM Data

Denoising Land-based Controlled-source Electromagnetic Data Based on a Same-site Noise Reference Channel

Learning a Model-Based Deep Hyperspectral Denoiser from a Single Noisy Hyperspectral Image.

Deep Learning for Unsupervised Denoising of Radar Sounder Data

Noise Modeling and Learning-Based Hyperspectral Image Denoising Used for Hyperspectral Unmixing

Single Image Super-Resolution Restoration of TGO CaSSIS Colour Images: Demonstration with Perseverance Rover Landing Site and Mars Science Targets

Denoising Hyperspectral Patches Between Manzius U & Boguslawsky M Lunar Craters from the Ch-1 M3 & Ch-2 IIRS Data

A hyperspectral image denoising method based on land cover spectral autocorrelation

Nonlocal Spatial–Spectral Neural Network for Hyperspectral Image Denoising

MaRF: Representing Mars as Neural Radiance Fields

Self-Supervised Elimination of Non-Independent Noise in Hyperspectral Imaging

Subpixel-Scale Topography Retrieval of Mars Using Single-Image DTM Estimation and Super-Resolution Restoration

Noise Performance of the Cris Instrument

DeepNRMS: Unsupervised Deep Learning for Noise-Robust CO2 Monitoring in Time-Lapse Seismic Images

Noise reduction in asteroid imaging using a miniaturized spectral imager

Detecting and Refining HiRISE Image Patches Obscured by Atmospheric Dust

Segmentation Convolutional Neural Networks for Automatic Crater Detection on Mars

Self-inspired learning for denoising live-cell super-resolution microscopy

Denoising of 3D Magnetic Resonance Images with Multi-Channel Residual Learning of Convolutional Neural Network