A neural network architecture for ultrasonic nondestructive testing

Y. Guez; K. D. Donohue; N. M. Bilgutay

doi:10.1109/ULTSYM.1991.234085

A neural network architecture for ultrasonic nondestructive testing

Y. Guez, K. D. Donohue, N. M. Bilgutay

Electrical and Computer Engineering

Research output: Contribution to journal › Conference article › peer-review

4 Scopus citations

Abstract

This paper considers the application of neural networks for detecting defects in materials consisting of dense micro-structures using ultrasonic pulse-echo systems. The motivation for this work is the desire to detect defects modeled by small complex scattering centers. To preserve signal features useful for defect detection, minimal preprocessing is performed on the data presented to the neural network. This paper describes modifications to the typical feed-forward multi-layer perceptron architecture for direct application to sampled RF ultrasonic A-scans.

Original language	English
Article number	234085
Pages (from-to)	777-780
Number of pages	4
Journal	Proceedings - IEEE Ultrasonics Symposium
DOIs	https://doi.org/10.1109/ULTSYM.1991.234085
State	Published - 1991
Event	1991 IEEE Ultrasonics Symposium. ULTSYM 1991 - Orlando, United States Duration: Dec 8 1991 → Dec 11 1991

Bibliographical note

Funding Information:
This work was partially supported by SDIO/IST managed by the Office of Naval Research under Contract No. N00014-86-K-0520 and by the National Science Foundation under Grant No. MIP-8920602.

Publisher Copyright:
© 1991 IEEE.

ASJC Scopus subject areas

Acoustics and Ultrasonics

Access to Document

10.1109/ULTSYM.1991.234085

Cite this

@article{4c56a9a8f2c14203b354d860a8e76253,

title = "A neural network architecture for ultrasonic nondestructive testing",

abstract = "This paper considers the application of neural networks for detecting defects in materials consisting of dense micro-structures using ultrasonic pulse-echo systems. The motivation for this work is the desire to detect defects modeled by small complex scattering centers. To preserve signal features useful for defect detection, minimal preprocessing is performed on the data presented to the neural network. This paper describes modifications to the typical feed-forward multi-layer perceptron architecture for direct application to sampled RF ultrasonic A-scans.",

author = "Y. Guez and Donohue, {K. D.} and Bilgutay, {N. M.}",

note = "Funding Information: This work was partially supported by SDIO/IST managed by the Office of Naval Research under Contract No. N00014-86-K-0520 and by the National Science Foundation under Grant No. MIP-8920602. Publisher Copyright: {\textcopyright} 1991 IEEE.; 1991 IEEE Ultrasonics Symposium. ULTSYM 1991 ; Conference date: 08-12-1991 Through 11-12-1991",

year = "1991",

doi = "10.1109/ULTSYM.1991.234085",

language = "English",

pages = "777--780",

}

TY - JOUR

T1 - A neural network architecture for ultrasonic nondestructive testing

AU - Guez, Y.

AU - Donohue, K. D.

AU - Bilgutay, N. M.

N1 - Funding Information: This work was partially supported by SDIO/IST managed by the Office of Naval Research under Contract No. N00014-86-K-0520 and by the National Science Foundation under Grant No. MIP-8920602. Publisher Copyright: © 1991 IEEE.

PY - 1991

Y1 - 1991

N2 - This paper considers the application of neural networks for detecting defects in materials consisting of dense micro-structures using ultrasonic pulse-echo systems. The motivation for this work is the desire to detect defects modeled by small complex scattering centers. To preserve signal features useful for defect detection, minimal preprocessing is performed on the data presented to the neural network. This paper describes modifications to the typical feed-forward multi-layer perceptron architecture for direct application to sampled RF ultrasonic A-scans.

AB - This paper considers the application of neural networks for detecting defects in materials consisting of dense micro-structures using ultrasonic pulse-echo systems. The motivation for this work is the desire to detect defects modeled by small complex scattering centers. To preserve signal features useful for defect detection, minimal preprocessing is performed on the data presented to the neural network. This paper describes modifications to the typical feed-forward multi-layer perceptron architecture for direct application to sampled RF ultrasonic A-scans.

UR - http://www.scopus.com/inward/record.url?scp=0013238151&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0013238151&partnerID=8YFLogxK

U2 - 10.1109/ULTSYM.1991.234085

DO - 10.1109/ULTSYM.1991.234085

M3 - Conference article

AN - SCOPUS:0013238151

SN - 1051-0117

SP - 777

EP - 780

JO - Proceedings - IEEE Ultrasonics Symposium

JF - Proceedings - IEEE Ultrasonics Symposium

M1 - 234085

T2 - 1991 IEEE Ultrasonics Symposium. ULTSYM 1991

Y2 - 8 December 1991 through 11 December 1991

ER -

A neural network architecture for ultrasonic nondestructive testing

Abstract

Bibliographical note

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this