2022
DOI: 10.1111/mice.12836
|View full text |Cite
|
Sign up to set email alerts
|

Automatic detection method of tunnel lining multi‐defects via an enhanced You Only Look Once network

Abstract: Aiming to solve the challenges of low detection accuracy, poor anti-interference ability, and slow detection speed in the traditional tunnel lining defect detection methods, a novel deep learning-based model, named You Only Look Once network v4 enhanced by EfficientNet and depthwise separable convolution (DSC; YOLOv4-ED), is proposed. In the YOLOv4-ED, EfficientNet is used as the backbone to improve the identification accuracy of indistinguishable defect targets in complex tunnel background and light condition… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
4
1

Citation Types

0
50
0

Year Published

2022
2022
2024
2024

Publication Types

Select...
7
1

Relationship

0
8

Authors

Journals

citations
Cited by 103 publications
(50 citation statements)
references
References 60 publications
0
50
0
Order By: Relevance
“…Given the failure mechanism of immersed tunnels, the key to the tunnel structure design lies in using accurate identification methods [18] and formulating effective prevention and control measures for standard joint failure modes. e shear key end angle, joint, and pipe segment connection should be optimized and strengthened.…”
Section: Damage and Failure Mechanism Of Tunnel And Design Key Pointsmentioning
confidence: 99%
“…Given the failure mechanism of immersed tunnels, the key to the tunnel structure design lies in using accurate identification methods [18] and formulating effective prevention and control measures for standard joint failure modes. e shear key end angle, joint, and pipe segment connection should be optimized and strengthened.…”
Section: Damage and Failure Mechanism Of Tunnel And Design Key Pointsmentioning
confidence: 99%
“…According to projections from the United Nations [5], if there is no change in how the construction sector operates, these impacts will be exacerbated by the scarcity of material resources, land use changes, and growing populations. In order to reduce the impacts of construction, the incorporation of waste as a raw material [6,7] or proposals to improve the mechanical properties of the material or the maintenance of buildings [8,9] are essential.…”
Section: Introductionmentioning
confidence: 99%
“…Actually, proximity detection can be referred to as obstacle recognition as well. If any vehicle/infrastructure in the vicinity can be inspected in advance, and the warning signals can be sent out accordingly, the successful rate of collision avoidance can be improved, and the false alarm problems can be also alleviated [ 5 , 6 ]. Decision-making subsystems consolidate the information obtained from the proximity detection subsystem to determine whether there is an imminent hazard.…”
Section: Introductionmentioning
confidence: 99%