A Hybrid Framework Employing Deep Learning for 3D Decay Segmentation and Adaptive Mapping of Heritage Structures: Insights from an Experiment

View Full Paper

Ask AI

Bookmark

Authors

Discussion

This framework uses machine learning to accurately detect decay types in tropical heritage structures, indicating improved diagnostic capabilities.

The unsupervised clustering algorithm achieved over 85% precision, outperforming traditional methods of decay detection.
Evaluations against ground-truth data emphasized accuracy and scalability in classifying chromatic variations of materials.
Application involves dual machine learning methods on RGB-enhanced photogrammetric data from heritage sites in Indonesia.
Findings support the framework's practical impact on enhancing digital heritage conservation workflows in humid environments.

Gbran et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark

This framework uses machine learning to accurately detect decay types in tropical heritage structures, indicating improved diagnostic capabilities.