MA-Xnet: Mobile-Attention X-Network for Crack Detection
Open Access
- 6 November 2022
- journal article
- research article
- Published by MDPI AG in Applied Sciences
- Vol. 12 (21) , 11240
Abstract
Modern crack detection algorithms based on deep learning have unsolved issues, such as an abundance of parameters in the resulting models and lack of context information. Such issues may lower the efficiency of feature extraction and lead to unexpected task performance. Based on two semantic segmentation models, U-Net and the dual-attention network (DANet), an efficient mobile-attention X-network (MA-Xnet) is proposed for crack detection. For performance evaluation, segmentation experiments were performed on concrete crack images from an internationally recognized dataset, which were collected from various campus buildings of Middle East Technical University. The experimental results demonstrated that, compared with U-Net, the proposed method parameters were reduced by 82.33%, and improved by 11.32% and 12.37% in the key indices of the F1-Score and the mean intersection of union (mIoU), respectively, providing a reference for subsequent related lightweight crack-segmentation research.Keywords
Funding Information
- National Natural Science Foundation of China (61873178, 201901D111093)
- the Natural Science Foundation of Shanxi (61873178, 201901D111093)
This publication has 31 references indexed in Scilit:
- Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation2018
- Automatic recognition of asphalt pavement cracks using metaheuristic optimized edge detection algorithms and convolution neural networkAutomation in Construction, 2018
- A Deep Convolutional Neural Network for Semantic Pixel-Wise Segmentation of Road and Pavement Surface Cracks2018
- Automatic Pixel‐Level Crack Detection and Measurement Using Fully Convolutional NetworkComputer-Aided Civil and Infrastructure Engineering, 2018
- Automatic crack inspection for concrete bridge bottom surfaces based on machine vision2017
- DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFsIEEE Transactions on Pattern Analysis and Machine Intelligence, 2017
- Road crack detection using deep convolutional neural network2016
- Hybrid ABC-ANN for pavement surface distress detection and classificationInternational Journal of Machine Learning and Cybernetics, 2015
- U-Net: Convolutional Networks for Biomedical Image Segmentation2015
- Fully convolutional networks for semantic segmentation2015