Extraction of line Surge Arresters from UAV LiDAR point clouds based on multi-view structural features
Bibliographic record
Abstract
• A structure-guided method identifies small Line Surge Arresters (LSAs) in 3D laser scans of power pylons. • Candidate line segments are selected using structural width and density estimation. • A swarm-optimization algorithm extracts LSAs via structural consistency measures. • The method achieves 98.28% identification accuracy and 90.11% F1-score on 3,571 pylons. • A public dataset of 896 annotated arresters supports 3D protection device research. Line Surge Arresters (LSAs) play a vital role in protecting power transmission systems from overvoltage, and obtaining their 3D information is crucial for precise reconstruction of power lines and intelligent planning of inspection routes. Current studies mainly rely on image-based methods for LSAs recognition. Unmanned Aerial Vehicle-mounted Light Detection and Ranging is an effective way to obtain 3D information of transmission corridors. However, accurately extracting LSAs from point clouds remains a significant challenge because of small physical size, sparse point distribution, and structural similarity to other components. In this study, an LSAs extraction method is proposed using the structural features. The method initially separates pylon and power line components using established techniques and further clusters the power line points into Single Power Line (SPL) as analysis unit, representing potential installation locations of the LSAs. For each SPL, this study proposes a structure-based method to identify the presence of LSAs. A width-based filtering criterion is applied to exclude SPLs without LSAs coarsely, retaining only those with potential LSA presence. For the retained SPLs, kernel density estimation is employed to capture the structural characteristics in case of LSA existence, thereby precisely preserving the SPLs containing LSAs. Following the identifying process, a Particle Swarm Optimization based segmentation method is proposed to achieve precise extraction of the LSAs. A structure-consistency-driven objective function is constructed using proposed Transmission Compactness Index and Axial Uniformity Index to model the geometric differences between LSAs and neighboring components. Based on this objective function, the optimal segmentation plane is searched to separate the LSAs from other components. The proposed method was evaluated on a dataset covering 63 transmission lines with a total of 3,571 pylons. Results demonstrate that the method achieves an overall identification accuracy of 98.28 %, with extraction precision of 92.66 %, recall of 87.43 %, and F1-score of 90.11 %. Additionally, our method shows high efficiency with the average processing time per pylon of 3.82 s. Compared to existing general segmentation algorithms, the proposed approach offers significant improvements in extraction accuracy. To foster further research, we release a point cloud dataset for LSAs extraction, which will be publicly available at: https://github.com/c175044/Line-Surge-Arresters-datasets .
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 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.003 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.002 |
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".