A frequency-texture-aware vision method for reliable measurement and inspection of concrete surface defects
Measurement Science and Technology · 13.11.2026 · 37 (45) · 455407
DOI: 10.1088/1361-6501/aeb8a9 ↗Kurz gefasst
Fachartikel in der Zeitschrift Measurement Science and Technology, online erschienen am 13.11.2026. Verfasst von Zheng-Gang Yang und 7 weiteren Autor:innen. Die Arbeit nennt 80 Literaturangaben. Gefördert von: Program of National Natural Science Foundation of China, Program of Chang Jiang Scholars of Ministry of Education, Liaoning Provincial Doctoral Research Start-up Fund Program.
Abstract (Originalsprache)
Vision-based concrete inspection estimates defect category, location, spatial extent and confidence from surface images, but repetitive texture, illumination variation, scale changes and feature misalignment undermine measurement reliability. We propose FTC-YOLO, a measurement-error-aware visual-inspection framework implemented with a frequency-texture-aware YOLOv8 detector. Its frequency-domain enhancement module re-weights spectral features to preserve defect boundaries and suppress low-frequency noise; the Texture-Adaptive Attention Module directs the neck to defect-bearing regions; and the cross-scale alignment fusion module aligns deep and shallow features before channel fusion. On a seven-category construction-defect benchmark, FTC-YOLO attains 0.920 ± 0.011 mAP@0.5 and 0.750 ± 0.013 mAP@0.5:0.95 across three seeds, versus 0.860 ± 0.012 and 0.685 ± 0.014 for YOLOv8s. On the common correctly matched ground-truth subset, the mean image-domain centre error decreases from 0.01764 (P95 0.02910) to 0.01224 (P95 0.02042). On a 400-image author-collected defect-free hold-out, false positives decrease from 0.10 to 0.07 per image and specificity increases from 0.90 to 0.93. On a disjoint 400-image author-collected positive hold-out with 2,486 annotations, FTC-YOLO achieves 0.905 ± 0.004 mAP@0.5 and 0.737 ± 0.006 mAP@0.5:0.95, versus 0.842 ± 0.004 and 0.663 ± 0.006 for YOLOv8s. The model retains at least 0.840 mAP@0.5 under brightness, blur, JPEG and input-size perturbations, and has 13.2 M parameters, requires 38.4 GFLOPs for a single forward pass, and achieves an inference speed of 128 FPS at an input resolution of 640 × 640 pixels on a single NVIDIA RTX 5090 GPU (batch size 1, FP16).
Förderung
- Program of National Natural Science Foundation of China
- Program of Chang Jiang Scholars of Ministry of Education
- Liaoning Provincial Doctoral Research Start-up Fund Program
- Science Discovery Award of Tencent Foundation
- Special Funds to Finance Operating Expenses for Basic Scientific Research of Central Colleges
Zitieren
APA
Yang, Z., Chen, Z., Wang, J., Liu, X., Hu, Z., Chen, J., He, S., & Li, A. (2026). A frequency-texture-aware vision method for reliable measurement and inspection of concrete surface defects. Measurement Science and Technology, 37(45), 455407. https://doi.org/10.1088/1361-6501/aeb8a9
Vancouver
Yang Z, Chen Z, Wang J, Liu X, Hu Z, Chen J, et al. A frequency-texture-aware vision method for reliable measurement and inspection of concrete surface defects. Measurement Science and Technology. 2026;37(45):455407. doi:10.1088/1361-6501/aeb8a9
BibTeX
@article{Yang2026,
author = {Yang, Zheng-Gang and Chen, Zheng-Han and Wang, Jiawei and Liu, Xiao-Jiang and Hu, Zhong-Wei and Chen, Jiong and He, Shengnan and Li, Ao-Li},
title = {A frequency-texture-aware vision method for reliable measurement and inspection of concrete surface defects},
journal = {Measurement Science and Technology},
year = {2026},
volume = {37},
number = {45},
pages = {455407},
doi = {10.1088/1361-6501/aeb8a9}
}
EUVIMED