EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Research on Synthetic Data Methods and Detection Models for Micro-Cracks

Yaotong Jiang, Tianmiao Wang, Xuanhe Chen, Jianhong Liang

Sensors · 17.03.2026 · 26 (6) · 1883

DOI: 10.3390/s26061883 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 17.03.2026. Verfasst von Yaotong Jiang und 3 weiteren Autor:innen. Beteiligte Einrichtung: School of Mechanical Engineering and Automation, Beihang University, Beijing 100191, China. Die Arbeit nennt 42 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Guangxi Key R & D Program Project "Intelligent Patrol and Defense System for Robots with Autonomous Control and Group Collaboration", Guangxi Key R & D Program Project “Intelligent Patrol and Defense System for Robots with Autonomous Control and Group Collaboration”.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Micro-crack detection on concrete surfaces is challenging because labeled micro-crack data are scarce, crack cues are extremely weak (often only a few pixels wide), and complex backgrounds (e.g., non-uniform illumination, shadows, and stains) degrade feature extraction; this study aims to improve both data availability and detection robustness for practical inspection. A Poisson image editing-based synthesis strategy is developed to generate visually coherent micro-crack samples via gradient-domain blending, and a Complex-Scene-Tolerant YOLO (CST-YOLO) detector is proposed on top of YOLOv10, following an “lighting decoupling–global perception–micro-feature enhancement” design. CST-YOLO integrates an Lighting-Adaptive Preprocessing Module (LAPM) to suppress illumination/shadow perturbations, a Spatial–Channel Sparse Transformer (SCS-Former) to model long-range crack topology efficiently, and a Small Object Focus Block (SOFB) to enhance micro-scale cues under cluttered backgrounds. Experiments are conducted on a 650-image dataset (200 real and 450 synthesized), in which synthesized samples are used only for training, and the validation/test sets contain only real images, with a 7:2:1 split. CST-YOLO achieves 0.990 mAP@0.5 and 0.926 mAP@0.5:0.95 at 139 FPS, and ablation results indicate complementary contributions from LAPM, SCS-Former, and SOFB. These results support the effectiveness of combining realistic synthesis and architecture-level robustness for real-time micro-crack detection in complex scenes.

Schlagwörter

Stichwörter der Autor:innen

Poisson image editingconcrete inspectionmicro-crack detectionreal-time detectionsmall-object enhancementsparse transformersynthetic data

Autor:innen und Institutionen

  • School of Mechanical Engineering and Automation, Beihang University, Beijing 100191, China.
  • School of Physical Science and Technology, Beijing University of Posts and Telecommunications, Beijing 102206, China.

Förderung

  • Guangxi Key R & D Program Project "Intelligent Patrol and Defense System for Robots with Autonomous Control and Group Collaboration"
  • Guangxi Key R & D Program Project “Intelligent Patrol and Defense System for Robots with Autonomous Control and Group Collaboration”

Zitieren

APA

Jiang, Y., Wang, T., Chen, X., & Liang, J. (2026). Research on Synthetic Data Methods and Detection Models for Micro-Cracks. Sensors, 26(6), 1883. https://doi.org/10.3390/s26061883

Vancouver

Jiang Y, Wang T, Chen X, Liang J. Research on Synthetic Data Methods and Detection Models for Micro-Cracks. Sensors. 2026;26(6):1883. doi:10.3390/s26061883

BibTeX
@article{Jiang2026,
  author = {Jiang, Yaotong and Wang, Tianmiao and Chen, Xuanhe and Liang, Jianhong},
  title = {Research on Synthetic Data Methods and Detection Models for Micro-Cracks},
  journal = {Sensors},
  year = {2026},
  volume = {26},
  number = {6},
  pages = {1883},
  doi = {10.3390/s26061883}
}
Kostenlos starten →