Papers for

satellite imaging teams

Papers whose findings have a practical use for this group, as judged from the abstract. Open a paper to read what it means in practice.

FastSAR speeds up satellite image making on CPUs GPUs and TPUs

FastSAR: High-Throughput, High-Fidelity Open-Source SAR Image Formation on CPUs, GPUs and TPUs

Abstract: Exact backprojection of a spaceborne SAR collection needs one interpolated read per pixel for every pulse; for the 12,207 by 8,808 pixel Umbra image of the Panama Canal this balloons to 1.6 trillion reads. FastSAR is an open-source library for SAR image formation on CPUs, NVIDIA GPUs and Google TPUs, with factorized backprojection, exact backprojection and polar format behind one Python interface and readers for the vendors' phase-history files. Its factorized kernels form the Panama image in 11.2 s on a 16-vCPU instance, 3.0 s on an NVIDIA L4, 0.6 s on an H100 and 2.9 s on a TPU v6e, with errors of -56 to -63 dB relative to a float64 exact backprojection in three regions over six cloud devices; the L4, a modest GPU, gives the lowest cost of the six at 0.59 US dollars per thousand images. At exact backprojection FastSAR is as fast as the MatX backprojection example, by far the fastest open-source implementation found (10.9 against 11.6 s at the same accuracy, with MatX's pixel coordinates in float64); its factorized algorithm is 3.8 times faster than MatX, and its polar format forms the image in 1.4 s on the L4 at -32 dB. On five Capella collections in three modes the worst-region error is below -59 dB.

Sat 24 OctPerformance
The gist
Creating detailed images from satellite radar data requires huge amounts of computing because each pixel depends on many measurements. The authors present FastSAR, an open-source software that makes this process much faster on different kinds of processors like CPUs, NVIDIA GPUs, and Google TPUs. FastSAR uses clever math methods to reduce the time needed to build very large images, making it practical to produce them in seconds rather than minutes or hours. It also matches or exceeds the accuracy of the best existing open tools.
Open → 2610.33932v1

Adaptive ensemble improves detail in solar magnetic field images

Super-Resolution of Solar Magnetograms via Adaptive Stratified Ensemble Learning with Uncertainty Estimation

Abstract: Single-image super-resolution of Sun's photospheric magnetograms enables consistent analysis across heterogeneous space-based instruments and supports long-term studies of solar magnetic field evolution. We address the super-resolution task from SOHO/MDI (low-resolution) to SDO/HMI (high-resolution) line-of-sight (LOS) magnetograms using a modified RRDBNet architecture initialized by ESRGAN pretrained weights. Through systematic per-image diagnostic analysis, we identify image complexity as the dominant predictor of reconstruction errors. To exploit this finding, we introduce an adaptive stratified specialist ensemble (SSE) of three specialist networks with uncertainty estimation, where each specialist network is trained by images from three different complexity strata using a weighted random sampling strategy. During inference, a lightweight router based on input image statistics assigns each test image to the appropriate specialist network. Our experimental results demonstrate the good performance of the proposed ensemble and its superiority over closely related methods.

Tue 22 SeptComputer Vision and Pattern Recognition
The gist
Images of the Sun's magnetic fields help scientists study solar activity, but different satellites produce pictures of varying detail. The authors created a computer method that turns lower-resolution images into higher-resolution ones similar to images from a better satellite. They found that the difficulty of each image affected how well this worked, so they made multiple specialized models to handle different image complexities. Their approach chooses the best model for each image, resulting in clearer and more accurate solar magnetic maps.
Open → 2609.27131v1

Radon transform enables precise alignment of space debris images

Sub-Pixel Affine Registration of Space Debris Images via the Radon Point Spread Function

Abstract: Inter-frame affine misalignment caused by platform jitter and attitude adjustments poses a fundamental challenge for multi-frame analysis of point targets in optical surveillance. Conventional registration methods rely on spatial intensity correlations or distinctive image features, both of which are largely absent in low-signal-to-noise-ratio point target imagery. We introduce the Radon Point Spread Function (RPSF) to characterize point targets in the Radon-transformed domain, and derive a closed-form framework that jointly estimates inter-frame translation and rotation from as few as four scalar RPSF samples per frame pair. The method requires no iterative optimization, feature extraction or interpolation, which is suitable for resource-constrained onboard processing. Simulation results confirm sub-pixel translation accuracy and a mean rotation error of 0.2556° at 1° Radon angular resolution. Validation on five real space debris datasets including both ground-based and in-orbit observations yields a mean calibration error below 0.5 pixels, substantially exceeding the precision required for reliable multi-frame processing.

Mon 7 SeptComputer Vision and Pattern RecognitionGraphics
The gist
Images of space debris taken from moving platforms often shift and rotate slightly, making it hard to analyze them together. The authors created a new way to measure these shifts and rotations very accurately by transforming the images using a mathematical tool called the Radon transform. Their method needs very little data, no guessing, and works well even when images are noisy. Tests with simulated and real images show their technique is better than what’s usually needed for combining multiple images.
Open → 2609.06929v1