EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Automated source detection in radio astronomy images using deep neural networks

Liang Cao, Chengkun Hou, Peng Jia, Nan Li, Huanyuan Shan, Lichun Sun, Zhenghao Zhu, Yangfan Xie, Heng Zhang, Jiameng Lv, Kaiyang Li, Mah Rukh und 1 weitere

Monthly Notices of the Royal Astronomical Society · 14.03.2026 · 547 (3)

DOI: 10.1093/mnras/stag376 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Monthly Notices of the Royal Astronomical Society, online erschienen am 25.02.2026. Verfasst von Liang Cao und 12 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 28 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China, National Key Research and Development Program of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Radio astronomical observations inherently capture both spatial and spectral emission, creating high-dimensional data cubes. As next-generation facilities like the Square Kilometre Array (SKA) come online, the volume and complexity of these data present significant challenges for automated source detection. Faint sources are often obscured by complex noise and artefacts, while their morphology evolves across the spectral range, limiting the efficacy of traditional threshold-based algorithms. To address these challenges, we present a volumetric detection methodology based on a 3D U-Net++ architecture designed to identify intricate structures within the 3D domain. We incorporate Monte Carlo dropout to perform Bayesian inference, providing robust per-detection uncertainty estimates. Validated against simulated data from the SKAO Science Data Challenge 2, our method demonstrates a substantial performance improvement over the standard Source Finding Application (SoFiA) pipeline, achieving a total score nearly 1000 points higher under the official metric. This gain is driven by superior sensitivity: our approach increases recall by approximately 2 per cent and recovers 22.2 per cent more confirmed sources than SoFiA, albeit with a 6 per cent decrease in precision. Furthermore, the uncertainty estimates derived from our Bayesian framework provide critical insights for candidate vetting and the prioritization of follow-up observations.

Förderung

  • National Natural Science Foundation of China
  • National Key Research and Development Program of China

Zitieren

APA

Cao, L., Hou, C., Jia, P., Li, N., Shan, H., Sun, L., Zhu, Z., Xie, Y., Zhang, H., Lv, J., Li, K., Rukh, M., & Omer, S. (2026). Automated source detection in radio astronomy images using deep neural networks. Monthly Notices of the Royal Astronomical Society, 547(3). https://doi.org/10.1093/mnras/stag376

Vancouver

Cao L, Hou C, Jia P, Li N, Shan H, Sun L, et al. Automated source detection in radio astronomy images using deep neural networks. Monthly Notices of the Royal Astronomical Society. 2026;547(3). doi:10.1093/mnras/stag376

BibTeX
@article{Cao2026,
  author = {Cao, Liang and Hou, Chengkun and Jia, Peng and Li, Nan and Shan, Huanyuan and Sun, Lichun and Zhu, Zhenghao and Xie, Yangfan and Zhang, Heng and Lv, Jiameng and Li, Kaiyang and Rukh, Mah and Omer, Sabiha},
  title = {Automated source detection in radio astronomy images using deep neural networks},
  journal = {Monthly Notices of the Royal Astronomical Society},
  year = {2026},
  volume = {547},
  number = {3},
  doi = {10.1093/mnras/stag376}
}
Kostenlos starten →