Pose-Informed X-ray Image Deblurring for Drone-Based Wind Turbine Inspection
e-Journal of Nondestructive Testing · 2026-07 · 31 (7)
DOI: 10.58286/33319 ↗Kurz gefasst
Fachartikel in der Zeitschrift e-Journal of Nondestructive Testing, 2026-07. Verfasst von Bas Meere und 5 weiteren Autor:innen.
Abstract (Originalsprache)
Motion blur presents a key challenge for drone-based X-ray inspection of wind turbine blades, as vibrations and drift severely degrade image quality. To address this, we introduce a geometric model that explicitly relates the relative motion of an independent source, object, and detector to the non-uniform blur in X-ray images. The model abstracts the imaging process into a lightweight, differentiable formulation suitable for use in optimization-based deblurring. By linking image formation directly to system motion, the model enables the direct use of pose data from on-board motion sensors to guide the deblurring process. Performance is demonstrated on both experimentally acquired and synthetically generated motion-blurred X-ray images of wind turbine blades. The results show that the model accurately reproduces complex blur patterns and forms a foundation for pose-informed motion deblurring in non-destructive testing.
Zitieren
APA
Meere, B., Heijne, R., Bondarev, E., Torta, E., Chanfreut, P., & Antunes, D. (2026). Pose-Informed X-ray Image Deblurring for Drone-Based Wind Turbine Inspection. e-Journal of Nondestructive Testing, 31(7). https://doi.org/10.58286/33319
Vancouver
Meere B, Heijne R, Bondarev E, Torta E, Chanfreut P, Antunes D. Pose-Informed X-ray Image Deblurring for Drone-Based Wind Turbine Inspection. e-Journal of Nondestructive Testing. 2026;31(7). doi:10.58286/33319
BibTeX
@article{Meere2026,
author = {Meere, Bas and Heijne, Rainier and Bondarev, Egor and Torta, Elena and Chanfreut, Paula and Antunes, Duarte},
title = {Pose-Informed X-ray Image Deblurring for Drone-Based Wind Turbine Inspection},
journal = {e-Journal of Nondestructive Testing},
year = {2026},
volume = {31},
number = {7},
doi = {10.58286/33319}
}
EUVIMED