A comparison of deep convolutional neural networks for image-based detection of concrete surface cracks
Authors:
- Marek Słoński
Abstract
The aim of this paper is to compare the performance of four deep convolutional neural networks in the problem of image-based automated detection of concrete surface cracks in the case of a small dataset. This crack detection problem is treated as a binary classification problem and it is solved by training a deep convolutional neural network on the small dataset. In this context, overfitting during training was the main issue to cope with and various techniques were applied to overcome this issue. The results of the experiments suggest that the best approach for this problem is to use the pretrained convolutional base of a large pretrained convolutional neural network as an automatic features extraction method and adding a new binary classifier on top of the convolutional base. Then training the new classifier and fine-tuning the last few layers of the pretrained network at the same time. The classification accuracy of the best deep convolutional neural network on the testing set is about 94 %.
- Record ID
- CUTa7d94b43f8ca4dcabe719f69a194a861
- Publication categories
- ;
- Author
- Journal series
- Computer Assisted Methods in Engineering and Science, ISSN 2299-3649
- Issue year
- 2019
- Vol
- 26
- No
- 2
- Pages
- 105-112
- Other elements of collation
- rys.; tab.; wykr.; Bibliografia (na s.) - 112; Bibliografia (liczba pozycji) - 22; Oznaczenie streszczenia - Streszcz. ang.; Numeracja w czasopiśmie - Vol. 26, No 2
- Substantive notes
- IPM 2019
- Keywords in English
- deep convolutional neural network, pretrained network, image based crack detection, binary classification, overfitting
- DOI
- DOI:10.24423/cames.267 Opening in a new tab
- URL
- https://cames.ippt.pan.pl/index.php/cames/article/view/267 Opening in a new tab
- Language
- eng (en) English
- License
- Score (nominal)
- 70
- Uniform Resource Identifier
- https://cris.pk.edu.pl/info/article/CUTa7d94b43f8ca4dcabe719f69a194a861/
- URN
urn:pkr-prod:CUTa7d94b43f8ca4dcabe719f69a194a861
* presented citation count is obtained through Internet information analysis, and it is close to the number calculated by the Publish or PerishOpening in a new tab system.