Physics-driven end-to-end wavefront coding for resilient imaging against sensor damage clusters
Photonics Research · 01.09.2026 · 14 (9) · 3999
DOI: 10.1364/prj.598863 ↗Kurz gefasst
Fachartikel in der Zeitschrift Photonics Research, online erschienen am 20.08.2026. Verfasst von Xiaofei Yu und 3 weiteren Autor:innen. Die Arbeit nennt 32 Literaturangaben. Gefördert von: National Key Research and Development Program of China, Anhui Provincial Key Research and Development Project, National Natural Science Foundation of China.
Abstract (Originalsprache)
The reliability of photoelectronic imaging sensors is frequently compromised in extreme environments by large-scale clustered physical damage, leading to irreversible information loss that exceeds the restorative capabilities of traditional interpolation-based algorithms. Here, we propose a physics-driven end-to-end framework designed to achieve high-resilience and high-fidelity imaging through the synergistic optimization of an optical wavefront coding frontend and a neural reconstruction backend. Grounded in statistical optics, we derive a spectral orthogonality constraint to minimize the overlap between the power spectral density of the system’s point spread function and the bad-pixel mask, guiding the frontend to implement active spatial redundancy encoding. The backend, a cascaded U-Net architecture, performs joint information interpolation and deconvolution under non-blind priors. Simulations and experiments demonstrate exceptional resilience against 100 random 40×40 pixels damage clusters, accurately restoring complex semantic features and supporting high-fidelity full-color reconstruction using a single physical mask. This paradigm shifts from passive hardware dependency to active physical redundancy, providing a robust foundation for imaging in extreme conditions.
Förderung
- National Key Research and Development Program of China
- Anhui Provincial Key Research and Development Project
- National Natural Science Foundation of China
- Fundamental Research Funds for the Central Universities
Zitieren
APA
Yu, X., Zang, H., Lu, Y., & Wang, P. (2026). Physics-driven end-to-end wavefront coding for resilient imaging against sensor damage clusters. Photonics Research, 14(9), 3999. https://doi.org/10.1364/prj.598863
Vancouver
Yu X, Zang H, Lu Y, Wang P. Physics-driven end-to-end wavefront coding for resilient imaging against sensor damage clusters. Photonics Research. 2026;14(9):3999. doi:10.1364/prj.598863
BibTeX
@article{Yu2026,
author = {Yu, Xiaofei and Zang, Haofeng and Lu, Yonghua and Wang, Pei},
title = {Physics-driven end-to-end wavefront coding for resilient imaging against sensor damage clusters},
journal = {Photonics Research},
year = {2026},
volume = {14},
number = {9},
pages = {3999},
doi = {10.1364/prj.598863}
}
EUVIMED