Autonomous Bridge Crack Detection Using Deep Convolutional Neural Networks

DOI

10.2991/iccia-19.2019.42How to use a DOI?

Keywords

Deep learning; convolution neural networks; bridge crack detection.

Abstract

Traditional image processing algorithms have a lot of limitations when dealing with crack detection problems. And the effect is not ideal if the classical deep learning model were used to detect bridge cracks directly. In order to solve these problems, a CNN-based bridge crack detection method is proposed in this paper, in which a feature extraction module based on arous space pyramid pool (ASPP) and depthwise separable convolution is designed. The former can obtain multi-scale image feature information, and the atrous convolution can provide a larger receptive field, so large-scale contextual information can be fused more effectively on feature maps. The latter can significantly reduce the computational complexity of the model and improve computational efficiency. The experimental results show that the method proposed in this paper achieved a crack detection accuracy of 96.69%, which is approximately 10% higher than other similar methods.

Copyright

© 2019, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Hongyan Xu
AU  - Xiu Su
AU  - Huaiyuan Xu
AU  - Haotian Li
PY  - 2019/07
DA  - 2019/07
TI  - Autonomous Bridge Crack Detection Using Deep Convolutional Neural Networks
BT  - Proceedings of the 3rd International Conference on Computer Engineering, Information Science & Application Technology (ICCIA 2019)
PB  - Atlantis Press
SP  - 274
EP  - 284
SN  - 2352-538X
UR  - https://doi.org/10.2991/iccia-19.2019.42
DO  - 10.2991/iccia-19.2019.42
ID  - Xu2019/07
ER  -