EUVIMEDEUVIMED entdecken
Crossref

Multi-view fusion and distillation for subgrade distresses detection based on 3D ground-penetrating radar

Chunpeng Zhou, Kangjie Ning, Haishuai Wang, Zhi Yu, Sheng Zhou, Jiajun Bu

Big Data Mining and Analytics · 20.11.2026

DOI: 10.26599/bdma.2026.9020039 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Big Data Mining and Analytics, online erschienen am 20.11.2026. Verfasst von Chunpeng Zhou und 5 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The Three-Dimensional Ground-Penetrating Radar (3D-GPR) has been widely used for subgrade distress detection. To enhance the efficiency and accuracy of detection, pioneering studies have attempted to adopt automatic detection techniques such as deep learning. However, existing works typically rely on traditional One-Dimensional (1D) A-scan, Two-Dimensional (2D)-B-scan or 3D C-scan GPR data, resulting in either insufficient spatial information or high computational complexity. To address these challenges, we introduce a novel methodology for subgrade distress detection by leveraging multi-view information from vanilla 3D-GPR data. First, we construct a real multi-view image dataset derived from the vanilla 3D-GPR data for the detection task, providing richer spatial information compared to A-scan and B-scan data, and reducing data computational complexity compared to C-scan data. Subsequently, we develop a novel Multi-View Fusion and Distillation Ground-Penetrating Radar framework, called GPR-MVFD, specifically designed to optimally utilize the multi-view GPR dataset. This framework ingeniously incorporates the attention-based fusion and multi-view distillation to facilitate significant feature extraction for subgrade distresses. In addition, a self-adaptive learning mechanism is adopted to stabilize the model training and prevent performance degeneration in each branch. Extensive experiments demonstrate the value of the newly constructed GPR dataset and showcase the effectiveness and efficiency of our proposed GPR-MVFD. Notably, our framework outperforms not only the existing GPR baselines, but also the state-of-the-art methods in the fields of multi-view learning, multimodal learning, and Knowledge Distillation (KD). Our data and code have been made publicly available as open source at https://github.com/zhouchunpong/multi-view-3DGPR.

Zitieren

APA

Zhou, C., Ning, K., Wang, H., Yu, Z., Zhou, S., & Bu, J. (2026). Multi-view fusion and distillation for subgrade distresses detection based on 3D ground-penetrating radar. Big Data Mining and Analytics. https://doi.org/10.26599/bdma.2026.9020039

Vancouver

Zhou C, Ning K, Wang H, Yu Z, Zhou S, Bu J. Multi-view fusion and distillation for subgrade distresses detection based on 3D ground-penetrating radar. Big Data Mining and Analytics. 2026. doi:10.26599/bdma.2026.9020039

BibTeX
@article{Zhou2026,
  author = {Zhou, Chunpeng and Ning, Kangjie and Wang, Haishuai and Yu, Zhi and Zhou, Sheng and Bu, Jiajun},
  title = {Multi-view fusion and distillation for subgrade distresses detection based on 3D ground-penetrating radar},
  journal = {Big Data Mining and Analytics},
  year = {2026},
  doi = {10.26599/bdma.2026.9020039}
}
Kostenlos starten →