Deep Learning Technology for Weld Defects Classification Based on Transfer Learning and Activation Features

Weld defects detection using X-ray images is an effective method of nondestructive testing. Conventionally, this work is based on qualified human experts, although it requires their personal intervention for the extraction and classification of heterogeneity. Many approaches have been done using mac...

Full description

Saved in:
Bibliographic Details
Main Authors: Chiraz Ajmi, Juan Zapata, Sabra Elferchichi, Abderrahmen Zaafouri, Kaouther Laabidi
Format: Article
Language:English
Published: Wiley 2020-01-01
Series:Advances in Materials Science and Engineering
Online Access:http://dx.doi.org/10.1155/2020/1574350
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849691445711077376
author Chiraz Ajmi
Juan Zapata
Sabra Elferchichi
Abderrahmen Zaafouri
Kaouther Laabidi
author_facet Chiraz Ajmi
Juan Zapata
Sabra Elferchichi
Abderrahmen Zaafouri
Kaouther Laabidi
author_sort Chiraz Ajmi
collection DOAJ
description Weld defects detection using X-ray images is an effective method of nondestructive testing. Conventionally, this work is based on qualified human experts, although it requires their personal intervention for the extraction and classification of heterogeneity. Many approaches have been done using machine learning (ML) and image processing tools to solve those tasks. Although the detection and classification have been enhanced with regard to the problems of low contrast and poor quality, their result is still unsatisfying. Unlike the previous research based on ML, this paper proposes a novel classification method based on deep learning network. In this work, an original approach based on the use of the pretrained network AlexNet architecture aims at the classification of the shortcomings of welds and the increase of the correct recognition in our dataset. Transfer learning is used as methodology with the pretrained AlexNet model. For deep learning applications, a large amount of X-ray images is required, but there are few datasets of pipeline welding defects. For this, we have enhanced our dataset focusing on two types of defects and augmented using data augmentation (random image transformations over data such as translation and reflection). Finally, a fine-tuning technique is applied to classify the welding images and is compared to the deep convolutional activation features (DCFA) and several pretrained DCNN models, namely, VGG-16, VGG-19, ResNet50, ResNet101, and GoogLeNet. The main objective of this work is to explore the capacity of AlexNet and different pretrained architecture with transfer learning for the classification of X-ray images. The accuracy achieved with our model is thoroughly presented. The experimental results obtained on the weld dataset with our proposed model are validated using GDXray database. The results obtained also in the validation test set are compared to the others offered by DCNN models, which show a best performance in less time. This can be seen as evidence of the strength of our proposed classification model.
format Article
id doaj-art-e89860ce1932486a80c7db1b48ce8364
institution DOAJ
issn 1687-8434
1687-8442
language English
publishDate 2020-01-01
publisher Wiley
record_format Article
series Advances in Materials Science and Engineering
spelling doaj-art-e89860ce1932486a80c7db1b48ce83642025-08-20T03:21:02ZengWileyAdvances in Materials Science and Engineering1687-84341687-84422020-01-01202010.1155/2020/15743501574350Deep Learning Technology for Weld Defects Classification Based on Transfer Learning and Activation FeaturesChiraz Ajmi0Juan Zapata1Sabra Elferchichi2Abderrahmen Zaafouri3Kaouther Laabidi4University of Tunis, National Superior Engineering of Tunis, Street Taha Hussein, Tunis, TunisiaUniversidad Politécnica de Cartagena, Departamento Tecnologías de la Informacion y las Comunicaciones, Campus la Muralla, Edif. Antigones 30202, Cartagena, Murcia, SpainUniversity Tunis El Manar, National Engineering School of Tunis, Analysis Design and Control of Systems Laboratory (LR11ES20), Tunis 1002, TunisiaUniversity of Tunis, National Superior Engineering of Tunis, Street Taha Hussein, Tunis, TunisiaUniversity of Jeddah, CEN Department, Jeddah, Saudi ArabiaWeld defects detection using X-ray images is an effective method of nondestructive testing. Conventionally, this work is based on qualified human experts, although it requires their personal intervention for the extraction and classification of heterogeneity. Many approaches have been done using machine learning (ML) and image processing tools to solve those tasks. Although the detection and classification have been enhanced with regard to the problems of low contrast and poor quality, their result is still unsatisfying. Unlike the previous research based on ML, this paper proposes a novel classification method based on deep learning network. In this work, an original approach based on the use of the pretrained network AlexNet architecture aims at the classification of the shortcomings of welds and the increase of the correct recognition in our dataset. Transfer learning is used as methodology with the pretrained AlexNet model. For deep learning applications, a large amount of X-ray images is required, but there are few datasets of pipeline welding defects. For this, we have enhanced our dataset focusing on two types of defects and augmented using data augmentation (random image transformations over data such as translation and reflection). Finally, a fine-tuning technique is applied to classify the welding images and is compared to the deep convolutional activation features (DCFA) and several pretrained DCNN models, namely, VGG-16, VGG-19, ResNet50, ResNet101, and GoogLeNet. The main objective of this work is to explore the capacity of AlexNet and different pretrained architecture with transfer learning for the classification of X-ray images. The accuracy achieved with our model is thoroughly presented. The experimental results obtained on the weld dataset with our proposed model are validated using GDXray database. The results obtained also in the validation test set are compared to the others offered by DCNN models, which show a best performance in less time. This can be seen as evidence of the strength of our proposed classification model.http://dx.doi.org/10.1155/2020/1574350
spellingShingle Chiraz Ajmi
Juan Zapata
Sabra Elferchichi
Abderrahmen Zaafouri
Kaouther Laabidi
Deep Learning Technology for Weld Defects Classification Based on Transfer Learning and Activation Features
Advances in Materials Science and Engineering
title Deep Learning Technology for Weld Defects Classification Based on Transfer Learning and Activation Features
title_full Deep Learning Technology for Weld Defects Classification Based on Transfer Learning and Activation Features
title_fullStr Deep Learning Technology for Weld Defects Classification Based on Transfer Learning and Activation Features
title_full_unstemmed Deep Learning Technology for Weld Defects Classification Based on Transfer Learning and Activation Features
title_short Deep Learning Technology for Weld Defects Classification Based on Transfer Learning and Activation Features
title_sort deep learning technology for weld defects classification based on transfer learning and activation features
url http://dx.doi.org/10.1155/2020/1574350
work_keys_str_mv AT chirazajmi deeplearningtechnologyforwelddefectsclassificationbasedontransferlearningandactivationfeatures
AT juanzapata deeplearningtechnologyforwelddefectsclassificationbasedontransferlearningandactivationfeatures
AT sabraelferchichi deeplearningtechnologyforwelddefectsclassificationbasedontransferlearningandactivationfeatures
AT abderrahmenzaafouri deeplearningtechnologyforwelddefectsclassificationbasedontransferlearningandactivationfeatures
AT kaoutherlaabidi deeplearningtechnologyforwelddefectsclassificationbasedontransferlearningandactivationfeatures