Deep Learning-Based Spatial Coherence Estimation for Ultrasonic NDT
e-Journal of Nondestructive Testing · 2026-07 · 31 (7)
DOI: 10.58286/33471 ↗Kurz gefasst
Fachartikel in der Zeitschrift e-Journal of Nondestructive Testing, 2026-07. Verfasst von Gabriel o. Vasconcellos und 3 weiteren Autor:innen.
Abstract (Originalsprache)
In this work, we adapt the CohereNet framework for Non-Destructive Testing (NDT) ultrasonic inspections using a compact DNN (Deep Neural Network) to estimate short-lag spatial coherence. Signals from three specimens acquired using plane-wave imaging were evaluated and compared against DAS (Delay And Sum) and traditional SLSC (Short-Lag Spatial Coherence) methods. Image quality was assessed using standard metrics, and a multi-criteria AHP–TOPSIS analysis was applied. Results show that the proposed SLSC-DNN achieves performance close to SLSC while reducing computational cost, indicating its potential as an efficient alternative for coherence-based NDT imaging.
Zitieren
APA
Vasconcellos, G. O., Dário, M. F., Prado, T. D. A., & Pipa, D. R. (2026). Deep Learning-Based Spatial Coherence Estimation for Ultrasonic NDT. e-Journal of Nondestructive Testing, 31(7). https://doi.org/10.58286/33471
Vancouver
Vasconcellos GO, Dário MF, Prado TDA, Pipa DR. Deep Learning-Based Spatial Coherence Estimation for Ultrasonic NDT. e-Journal of Nondestructive Testing. 2026;31(7). doi:10.58286/33471
BibTeX
@article{Vasconcellos2026,
author = {Vasconcellos, Gabriel o. and Dário, Matheus F. and Prado, Tatiana de Almeida and Pipa, Daniel Rodrigues},
title = {Deep Learning-Based Spatial Coherence Estimation for Ultrasonic NDT},
journal = {e-Journal of Nondestructive Testing},
year = {2026},
volume = {31},
number = {7},
doi = {10.58286/33471}
}
EUVIMED