Quasi-online three-dimensional reconstruction of worn topography using a low-cost kaleidoscopic imaging system
Applied Optics · 01.08.2026 · 65 (22) · 7664
DOI: 10.1364/ao.601596 ↗Kurz gefasst
Fachartikel in der Zeitschrift Applied Optics, online erschienen am 30.07.2026. Verfasst von Yinhu Xi und 2 weiteren Autor:innen. Die Arbeit nennt 31 Literaturangaben. Gefördert von: Anhui Key Laboratory of Mine Intelligent Equipment and Technology, Anhui University of Science and Technology.
Abstract (Originalsprache)
Worn surface topography is vital for revealing wear mechanisms. However, traditional in situ measurement techniques are limiting dynamic monitoring. To overcome this limitation, this study proposes a quasi-online 3D reconstruction method integrating a kaleidoscopic imaging system with the DUSt3R neural network. This approach captures multi-view image sequences during the wear process and employs the neural network to generate 3D topography. Experimental results demonstrate that, compared to the ground -truth obtained by a 3D surface profilometer, the average relative errors of wear depth and width parameters are 11.33% and 12.27%, respectively. This study provides an effective technical solution for continuous monitoring of wear morphology without equipment shutdown.
Förderung
- Anhui Key Laboratory of Mine Intelligent Equipment and Technology, Anhui University of Science and Technology
Zitieren
APA
Xi, Y., Liang, S., & Fei, S. (2026). Quasi-online three-dimensional reconstruction of worn topography using a low-cost kaleidoscopic imaging system. Applied Optics, 65(22), 7664. https://doi.org/10.1364/ao.601596
Vancouver
Xi Y, Liang S, Fei S. Quasi-online three-dimensional reconstruction of worn topography using a low-cost kaleidoscopic imaging system. Applied Optics. 2026;65(22):7664. doi:10.1364/ao.601596
BibTeX
@article{Xi2026,
author = {Xi, Yinhu and Liang, Shenao and Fei, Shuhui},
title = {Quasi-online three-dimensional reconstruction of worn topography using a low-cost kaleidoscopic imaging system},
journal = {Applied Optics},
year = {2026},
volume = {65},
number = {22},
pages = {7664},
doi = {10.1364/ao.601596}
}
EUVIMED