| Title |
Development of Tunnel Lining Crack Detection Device and Crack Detection for Smart Construction diagnosis |
| Authors |
김우영(Kim, Woo-Young) ; 배재훈(Bae, Jae-hoon) |
| DOI |
https://doi.org/10.5659/JAIK.2024.40.11.225 |
| Keywords |
Concrete Crack; Crack Detection; Deep Learning; Tunnel Construction diagnosis; Smart Construction |
| Abstract |
The purpose of this study is to develop a tunnel lining crack detection device mounted on an automatic tunnel inspection robot for smart
construction diagnosis and to examine a crack detection method using deep learning. The tunnel lining crack detection device is modularized
by combining a Pan-Tilt device, a green monochrome pulse irradiation device, a LiDAR range finder, and an automatic zoom adjustment
device. To verify the crack detection model using deep learning, a crack simulation specimen was produced, and YOLOv5 was selected as
the crack detection model. The crack detection model evaluation results showed a performance of recall 97.4%, precision 97.6%, and
mAP50:95 80.4%. In addition, a crack simulation specimen was photographed 6m away from the crack detection device, and it was
confirmed that all cracks in the photographed crack simulation specimen images were detected. |