Residual diffusion for fine-structure enhancement in low-count accelerator-based muon imaging

Accelerator-based muon imaging combines the penetrating power of high-energy muons with selectable momentum and a controlled beam direction, offering a powerful means of probing internal structures in large, dense and heavily shielded objects. Within a fixed beam-time budget, realizing this potential requires reliable recovery of fine structures from limited muon statistics. We propose an observation-conditioned residual diffusion method with decoupled linear schedules to enhance low-count muon images through iterative restoration guided by the observed image. We evaluate a simplified VSC-24 fuel cask under an idealized 5 GeV parallel muon beam, using 500,000 accepted events per low-count image and 40 million per reference image. Simulation results show improved recovery of structural boundaries and fine details, with the mean peak signal-to-noise ratio increasing from 11.4239 to 22.1719 dB. These findings highlight the potential of diffusion models to make more effective use of limited muon data and broaden the application of accelerator-based muon imaging across scientific and industrial settings.

Authors

Publication Details

Journal
arXiv (Cornell University)
Published
2026-11-18
DOI
https://doi.org/10.48550/arxiv.2611.23776
Primary Topic
Particle Detector Development and Performance
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Residual diffusion for fine-structure enhancement in low-count accelerator-based muon imaging

Zhengyang He, Pei Yu, Zhiyu Sun, Yuhong Yu et al.
arXiv (Cornell University)
Particle Detector Development and Performance
preprint

Residual diffusion for fine-structure enhancement in low-count accelerator-based muon imaging

Zhengyang He, Pei Yu, Zhiyu Sun, Yuhong Yu, Xueheng Zhang, Yu Zhang, Liangwen Chen, Haochen Wang, Lei Yang
preprint en

Abstract

Accelerator-based muon imaging combines the penetrating power of high-energy muons with selectable momentum and a controlled beam direction, offering a powerful means of probing internal structures in large, dense and heavily shielded objects. Within a fixed beam-time budget, realizing this potential requires reliable recovery of fine structures from limited muon statistics. We propose an observation-conditioned residual diffusion method with decoupled linear schedules to enhance low-count muon images through iterative restoration guided by the observed image. We evaluate a simplified VSC-24 fuel cask under an idealized 5 GeV parallel muon beam, using 500,000 accepted events per low-count image and 40 million per reference image. Simulation results show improved recovery of structural boundaries and fine details, with the mean peak signal-to-noise ratio increasing from 11.4239 to 22.1719 dB. These findings highlight the potential of diffusion models to make more effective use of limited muon data and broaden the application of accelerator-based muon imaging across scientific and industrial settings.

arXiv (Cornell University)
Particle Detector Development and Performance
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon