.png){kind=icon}
.png){kind=icon}
Return
Pixel-level classification for bridge deck rebar detection and localization using multi-stage deep encoder-decoder network
.png){kind=icon}
OA
AI
DOI：10.1016/j.dibe.2023.100132.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Structural Health Monitoring (SHM) and Nondestructive Evaluation (NDE) of civil infrastructure have been active areas of research for the past few decades. The traditional inspection methods for civil infrastructure, mostly relying on the visual inspection are time-consuming, labor-intensive, error-prone, and often provide subjective results. In the wake of rising costs for infrastructural maintenance, the time factor, safety issues, and the error-prone nature of human inspection methods, there is an increased need for the development of auto-mated methods for bridge inspection and maintenance. The purpose of this research is to provide a novel Deep Learning-based approach for rebar detection and localization within bridge decks. The proposed system is trained using Ground Penetrating Radar (GPR) data from 8 real bridges in the United States. The results have been discussed in terms of qualitative and quantitative aspects with considerable potential and various issues that need to be explored in future works. Due to the similarity in the type of parabolic signatures present in other GPR-related applications, this technique can be generalizable to other applications. The proposed approach for rebar detection and localization has considerable implications for the civil experts in general and GPR com-munity in particular.
Keywords:
Structural health monitoring (SHM)
Non-destructive evaluation
Ground penetrating radar (GPR)
Rebar detection and localization
Deep convolutional neural networks
Deep encoder-decoder networks
Pixel-Level classification
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
8.2
Papers:
998
Citations:
3.3K