ASTRA-SR: Atmospheric Seeing and Turbulence Restoration for Astronomical Image Super-Resolution

Xining Ge; Ziteng Cui; Shuhong Liu

cs.CVSep 22, 2026

ASTRA-SR: Atmospheric Seeing and Turbulence Restoration for Astronomical Image Super-Resolution

Authors: Xining Ge, Ziteng Cui, Shuhong Liu

Organizations: Hangzhou Dianzi University · Hong Kong University of Science and Technology, Guangzhou · The University of Tokyo

Abstract

Ground-based planetary imaging suffers from atmospheric turbulence, sensor noise, and limited sampling, making restoration a joint denoising, deblurring, and super-resolution problem. We present ASTRA-SR, a blind single-frame restoration framework trained on a physics-grounded synthetic dataset. High-dynamic-range spacecraft RAW observations serve as clean sources, and paired LR inputs are synthesized using measured layer-integrated turbulence strengths, propagated moving phase screens, exposure-averaged spatially varying PSFs, and sensor noise.ASTRA-SR first estimates a noise-suppressed but blur-retaining LR image, then restores spatial structure through multiscale processing and reconstructs HR detail with serial spatial-amplitude refinement. It yields a 0.49 dB foreground PSNR gain over the strongest baseline approaches.

Figures & tables

Fig. 1: ASTRA-SR architecture. The synthesis path generates the degraded LR input and training-only targets; the restoration path contains the three-scale SCGN backbone, intermediate LR estimation branch, stage-specific refinement, and sequential spatial-amplitude HR head.

Fig. 2: Representative PSFs and layer-strength sampling models for an illustrated MASS profile.

Fig. 3: Fourier-component restoration analysis under full, PSF-only, and noise-only degradation.

Method	Venue	PSNR $\uparrow$	SSIM $\uparrow$	Obj.-PSNR $\uparrow$
Bicubic [ 14 ]	–	29.868	0.778	25.547
Bilinear [ 16 ]	–	29.757	0.776	25.427
NAFNet [ 4 ]	ECCV’22	35.275	0.843	31.434
Restormer [ 40 ]	CVPR’22	32.615	0.819	28.602
FFTformer [ 15 ]	CVPR’23	35.131	0.844	31.547
SMFANet [ 44 ]	ECCV’24	33.605	0.836	29.605

Tab. 1 : Quantitative comparison on 1,355 synthetic test images.

Fig. 4: Qualitative comparisons on paired synthetic data and real observations. All methods use identical coordinates and zoom regions. The real Jupiter and lunar observations have no reference targets.

Dimension	Variant	PSNR	$\Delta$ (dB)
Full model	ASTRA-SR full model	35.824	–
LR estimation	w/o frequency-band adapter	35.734	0.090
	w/o intermediate LR supervision	35.719	0.105
LR restoration	w/o dilated branches	35.713	0.112
	w/o Patch-Fourier refiner	35.739	0.085
	w/o local residual gain	35.733	0.091

Tab. 2 : Matched ASTRA ablation by module dimension.

\Delta

is computed from unrounded PSNR values.

Explore similar work

AstraMoE-SR: Trajectory-Guided Diffusion for Blind Satellite Jitter Deblurring and Super-Resolution
Sep 7, 2026Yi-Chung Lai, Chin-Tien Wu, Yu-Chih ChenImage Deblurring Remote Sensing Image Super-Resolution
Institute of Mathematical Modeling and Scientific Computing National Yang Ming Chiao Tung University · Department of Computer Science National Yang Ming Chiao Tung University
FluxFlow: Conservative Flow-Matching for Astronomical Image Super-Resolution
May 5, 2026Shuhong Liu, Xining Ge, Ziteng Cui +8Flow Matching Image Super-Resolution
1The University of Tokyo · 2I2WM · 3Tohoku University +1
GSTurb: Gaussian Splatting for Atmospheric Turbulence Mitigation
Feb 26, 2026Hanliang Du, Zhangji Lu, Zewei Cai +3Image Restoration Image Deblurring
Faculty of Science and Technology, University of Macau, Macau 999078, China · College of Physics and Optoelectronic Engineering, Shenzhen University, Shenzhen 518000, China · College of Physics and Optoelectronic Engineering, Shenzhen University, Shenzhen, 518060, China +4

Sep 7, 2026J/K move · Enter open · S save

AstraMoE-SR: Trajectory-Guided Diffusion for Blind Satellite Jitter Deblurring and Super-Resolution

Yi-Chung Lai, Chin-Tien Wu, Yu-Chih Chen

Institute of Mathematical Modeling and Scientific Computing National Yang Ming Chiao Tung University · Department of Computer Science National Yang Ming Chiao Tung University

Image Deblurring Remote Sensing Image Super-Resolution Diffusion Models for Image Restoration

Open paper AskPDF arXiv

Pushbroom satellite imaging couples limited spatial resolution with platform attitude instability. Platform jitter produces spatially varying motion blur because each scan line is acquired under a different instantaneous attitude, while perspective geometry causes the same perturbation to induce different pixel displacements across the field of view. Existing blind restoration methods that assume a spatially invariant kernel and satellite jitter correction methods that rely on auxiliary observations are therefore not directly applicable. We present AstraMoE-SR, a single-image framework that jointly restores motion blur and spatial resolution without auxiliary measurements. Rather than estimating a blur kernel, we infer how the camera moved by reparameterizing degradation as a local exposure trajectory under pushbroom geometry. A conditional diffusion model estimates the trajectory distribution, mitigating the over-smoothing of high-frequency jitter by deterministic point estimation. The predicted trajectory conditions a pretrained latent diffusion backbone through trajectory-guided geometric alignment and spatially adaptive reconstruction. We further show that the remaining point-wise trajectory error is consistent with intrinsic jitter-phase ambiguity that is not resolved by increasing estimator capacity. On all 1,411 DOTA-v1.0 images degraded using our physically motivated forward model, AstraMoE-SR is the only evaluated method to outperform the no-restoration baseline across every fidelity metric, improving on StableSR by 0.64 dB PSNR, 15.2% LPIPS, and 0.091 DINO feature similarity. Reconstructions conditioned on predicted trajectories differ negligibly from those using ground-truth trajectories, indicating that the estimates retain the degradation information required for effective restoration.

Abstract

Explore similar work

AstraMoE-SR: Trajectory-Guided Diffusion for Blind Satellite Jitter Deblurring and Super-Resolution

FluxFlow: Conservative Flow-Matching for Astronomical Image Super-Resolution

GSTurb: Gaussian Splatting for Atmospheric Turbulence Mitigation