EUVIMEDEUVIMED entdecken
Crossref

Event-Guided Infrared Video Deblurring

Xintong Han, Jianing Li, Yunjian Zhang, Tianming Zhao, Shaoyu Liu, Luxin Yan, Xiangyang Ji

IEEE Transactions on Circuits and Systems for Video Technology · 2026-11 · 36 (11) · 17049-17063

DOI: 10.1109/tcsvt.2026.3721091 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Circuits and Systems for Video Technology, 2026-11. Verfasst von Xintong Han und 6 weiteren Autor:innen. Die Arbeit nennt 92 Literaturangaben. Gefördert von: National Natural Science Foundation of China, National Key Research and Development Program of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Infrared cameras are widely used in low-light and harsh weather conditions. However, motion blur can severely degrade both visual quality and the usefulness of infrared videos. In this paper, we propose a novel event-guided infrared video deblurring (EIVDeblur) model, exploring the use of high-temporal-resolution event streams for infrared video deblurring. Technically, we first establish a realistic synthetic dataset and a real-sensor event-infrared video dataset with synthetic blur generation for deblurring, where the blurred samples are generated synthetically while maintaining temporal association and approximate spatial alignment between event streams and infrared frames. Then, a dynamic dual-guided attention fusion module is designed to employ attention mechanisms to adaptively correlate motion cues from event streams with texture information in infrared images. Finally, a bidirectional gated recurrent module is presented to utilize rich temporal cues from two visual streams and enhance long-term dependency modeling. Extensive experiments on both datasets demonstrate that the proposed EIVDeblur model outperforms state-of-the-art methods in PSNR, SSIM, and LPIPS metrics. We believe that this hybrid imaging paradigm will provide new insights into future infrared camera systems. Our dataset and code can be available at https://github.com/xintong-han/event_guided_infrared_video_deblurring

Förderung

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

Zitieren

APA

Han, X., Li, J., Zhang, Y., Zhao, T., Liu, S., Yan, L., & Ji, X. (2026). Event-Guided Infrared Video Deblurring. IEEE Transactions on Circuits and Systems for Video Technology, 36(11), 17049-17063. https://doi.org/10.1109/tcsvt.2026.3721091

Vancouver

Han X, Li J, Zhang Y, Zhao T, Liu S, Yan L, et al. Event-Guided Infrared Video Deblurring. IEEE Transactions on Circuits and Systems for Video Technology. 2026;36(11):17049-17063. doi:10.1109/tcsvt.2026.3721091

BibTeX
@article{Han2026,
  author = {Han, Xintong and Li, Jianing and Zhang, Yunjian and Zhao, Tianming and Liu, Shaoyu and Yan, Luxin and Ji, Xiangyang},
  title = {Event-Guided Infrared Video Deblurring},
  journal = {IEEE Transactions on Circuits and Systems for Video Technology},
  year = {2026},
  volume = {36},
  number = {11},
  pages = {17049--17063},
  doi = {10.1109/tcsvt.2026.3721091}
}
Kostenlos starten →