Automated Bridge Inspection for Concrete Surface Defect Detection Using Deep Neural Network Based on LiDAR Scanning
Bibliographic record
Abstract
Structural inspection and maintenance of bridges are essential to improve the safety and sustainability of the infrastructure systems. Visual inspection using non-equipped eyes is the principal method of detecting surface defects of bridges, which is time-consuming, unsafe, and encounters inspectors falling risks. Therefore, there is a need for automated bridge inspection. Recently, Light Detection and Ranging (LiDAR) scanners are used for detecting surface defects. LiDAR scanners can collect high-quality 3D point cloud datasets. In order to automate the process of structural inspection, it is important to collect proper datasets and use an efficient approach to analyze them and find the defects. Deep Neural Networks (DNNs) have been recently used for detecting 3D objects within 3D point clouds. PointNet and PointNet++ are deep neural networks for classification, part segmentation, and semantic segmentation of point clouds that are modified and adapted in this work to detect surface concrete defects. The research contributions are: (1) Designing a LiDAR-equipped UAV platform for structural inspection using an affordable 2D scanner for data collection, and (2) Proposing a method for detecting concrete surface defects using deep neural networks based on LiDAR generated point clouds. Training and testing datasets are collected from four concrete bridges in Montréal and annotated manually. The point cloud dataset prepared in five areas, which contain more than 51 million points and 2,572 annotated defects in four classes of crack, light spalling, medium spalling, and severe spalling. The accuracies of 75% (adapted PointNet) and 79% (adapted PointNet++) in detecting defects are achieved in binary semantic segmentation.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".