A lightweight neural radiance field model with automatic semantic segmentation for post‐earthquake building three‐dimensional reconstruction

View Full Paper

Ask AI

Bookmark

LNAs achieved a peak signal-to-noise ratio of 26.20, indicating enhanced 3D reconstruction quality.
Automatic semantic segmentation in LNAs resulted in a mean intersection over union of 84.52%, optimizing detection accuracy.
Observational analysis utilizing UAV-captured images enables efficient 3D continuous modeling with semantic details.
Significant performance improvements suggest that LNAs could transform post-earthquake damage assessment strategies.

Discussion

No takes yet. Share an insight, caveat, or question.

This method demonstrates improved semantic segmentation and 3D reconstruction in buildings, suggesting rapid damage assessment with UAVs.

Luo et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark