Publication: Image-based concrete crack detection using enhanced convolutional neutral network
| datacite.subject.fos | oecd::Engineering and technology::Civil engineering | |
| dc.contributor.author | He, Mianqing | |
| dc.date.accessioned | 2025-11-27T03:52:33Z | |
| dc.date.available | 2025-11-27T03:52:33Z | |
| dc.date.issued | 2024-10-01 | |
| dc.description.abstract | The maintenance of concrete takes considerable time and poses a significant task, especially when it comes to detecting cracks at the pixel level. Manual visual inspection, while the most straightforward method for assessing concrete surface damage, is labor-intensive, time-consuming, incurs high labor costs, exhibits low efficiency, and yields subjective results with poor repeatability. Moreover, conducting inspections in hazardous areas poses significant challenges for humans. This research presents a novel methodology termed as CrackHAM, which is an encoder-decoder network founded on the U-Net architecture. The primary objectives of CrackHAM are twofold: to achieve accurate and robust concrete crack detection while reducing the parameters of the network. This study introduces two significant improvements to the existing neural network architecture, namely the phased multi-fusion module and the dual attention mechanisms. Furthermore, a novel module named HASPP is devised to augment the network’s capacity to acquire more comprehensive receptive fields. In order to lower the number of network parameters, a technique is employed whereby only use half of the number of input channels and output channels in the VGG16 are utilized as U-Net encoder modules. The empirical findings demonstrate that in the DeepCrack, Crack500, and FIND public datasets, CrackHAM achieves superior segmentation performance compared to the FCN, Deeplabv3, Swin-Unet, U-Net, and other models while utilizing only one-third of the computational resources. Quantitative evaluations show that CrackHAM achieves an IoU of 0.7724 on the DeepCrack dataset, 0.5973 on the Crack500 dataset, and 0.7644 on the FIND dataset. The network was applied on mobile device and can segment the crack accurately. | |
| dc.identifier.uri | https://erepo.usm.my/handle/123456789/23193 | |
| dc.language.iso | en | |
| dc.title | Image-based concrete crack detection using enhanced convolutional neutral network | |
| dc.type | Resource Types::text::thesis::master thesis | |
| dspace.entity.type | Publication | |
| oairecerif.author.affiliation | Universiti Sains Malaysia |