EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

An Adaptive Edge-Guided Dual-Network Framework for Fast QR Code Motion Deblurring

Jianping Li iD, Dongyang Guo, Wenjie Li, Wei Zhao

Sensors · 03.08.2026 · 26 (15) · 4879

DOI: 10.3390/s26154879 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 03.08.2026. Verfasst von Jianping Li und 3 weiteren Autor:innen. Beteiligte Einrichtung: Guangdong Provincial Key Lab of Robotics and Intelligent Systems, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen 51…. Die Arbeit nennt 48 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Guandong Key Laboratory for New Generation IoT Research and Development, Guangdong Key Laboratory for New Generation IoT Research and Development.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Unlike natural image deblurring, which primarily emphasizes perceptual quality and pixel-level fidelity, Quick Response (QR) code deblurring must retain decoding-critical structures to guarantee successful decoding. QR codes contain regular binary module grids and functional patterns with sharp boundaries, providing a strong structural prior for restoration. However, most existing learning-based QR restoration methods capture QR-specific structural information via implicit feature learning. To address this limitation, we propose an Edge-Guided Attention Block (EGAB), which explicitly extracts multi-directional edge priors and injects them into the query–key correlations of Transformer attention. Based on EGAB, we develop an Edge-Guided Restormer (EG-Restormer) for restoring severely blurred QR codes. For mildly blurred inputs, we introduce a Lightweight and Efficient Network (LENet) that performs fast restoration with low computational overhead. We further integrate EG-Restormer and LENet into an Adaptive Dual-network (ADNet), which selects the appropriate restoration branch according to the input blur level. Extensive experiments demonstrate the effectiveness of the proposed framework. EG-Restormer boosts the decoding rate by 8.67 percentage points under GoPro-only training and achieves the highest decoding rate among the evaluated methods after QRData fine-tuning. Moreover, ADNet reduces average inference latency by 19% while maintaining comparable decoding performance. These results suggest that explicit edge prior modeling enhances the recovery of structures critical for decoding, while adaptive routing provides an effective balance between decoding accuracy and computational efficiency.

Schlagwörter

Stichwörter der Autor:innen

QR code motion debluringadaptive dual networklightweight image deblurringmotion deblurringstructural image prior

Autor:innen und Institutionen

  • Guangdong Provincial Key Lab of Robotics and Intelligent Systems, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, China.
  • University of Chinese Academy of Sciences, Beijing 101408, China.
  • School of Nano Science and Technology, University of Science and Technology of China, Suzhou 215127, China.
  • School of Computer Science and Artificial Intelligence, Shenzhen University of Advanced Technology, Shenzhen 518107, China.

ORCID-Profile: Jianping Li

Förderung

  • Guandong Key Laboratory for New Generation IoT Research and Development
  • Guangdong Key Laboratory for New Generation IoT Research and Development

Zitieren

APA

Li, J., Guo, D., Li, W., & Zhao, W. (2026). An Adaptive Edge-Guided Dual-Network Framework for Fast QR Code Motion Deblurring. Sensors, 26(15), 4879. https://doi.org/10.3390/s26154879

Vancouver

Li J, Guo D, Li W, Zhao W. An Adaptive Edge-Guided Dual-Network Framework for Fast QR Code Motion Deblurring. Sensors. 2026;26(15):4879. doi:10.3390/s26154879

BibTeX
@article{Li2026,
  author = {Li, Jianping and Guo, Dongyang and Li, Wenjie and Zhao, Wei},
  title = {An Adaptive Edge-Guided Dual-Network Framework for Fast QR Code Motion Deblurring},
  journal = {Sensors},
  year = {2026},
  volume = {26},
  number = {15},
  pages = {4879},
  doi = {10.3390/s26154879}
}
Kostenlos starten →