Denoising the Deep Sky: Physics-Based CCD Noise Formation for Astronomical Imaging

2026-01-30

AI Summary

Astronomical imaging is severely constrained by complex noise characteristics and the absence of real paired data for supervised denoising. This work proposes a physics-based CCD noise synthesis framework that, for the first time, systematically models the full noise pipeline—including photon shot noise, pixel response non-uniformity, dark current, readout noise, and cosmic ray impacts—and generates high-fidelity paired training data by averaging multiple unregistered exposures to produce high-SNR ground-truth images. Leveraging real observations from two telescopes, the authors construct a realistic dataset spanning multiple bands and encompassing raw frames, calibrated frames, and stacked images. This dataset enables the training of interpretable, high-performance denoising models and establishes a standardized benchmark for evaluating astronomical image processing methods.

Technology Category

• Computer Vision: Computational Photography, Image & Video Synthesis
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Calibration & Uncertainty Quantification

Application Category

• Security and Privacy: Data transparency and provenance
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs

Abstract

Astronomical imaging remains noise-limited under practical observing constraints, while standard calibration pipelines mainly remove structured artifacts and leave stochastic noise largely unresolved. Learning-based denoising is promising, yet progress is hindered by scarce paired training data and the need for physically interpretable and reproducible models in scientific workflows. We propose a physics-based noise synthesis framework tailored to CCD noise formation. The pipeline models photon shot noise, photo-response non-uniformity, dark-current noise, readout effects, and localized outliers arising from cosmic-ray hits and hot pixels. To obtain low-noise inputs for synthesis, we average multiple unregistered exposures to produce high-SNR bases. Realistic noisy counterparts synthesized from these bases using our noise model enable the construction of abundant paired datasets for supervised learning. We further introduce a real-world dataset across multi-bands acquired with two twin ground-based telescopes, providing paired raw frames and instrument-pipeline calibrated frames, together with calibration data and stacked high-SNR bases for real-world evaluation.

Problem

• astronomical imaging
• CCD noise
• denoising
• paired training data
• stochastic noise

Innovation

• physics-based noise modeling
• CCD noise synthesis
• astronomical image denoising
• paired training data generation
• real-world astronomical dataset