About this document
Unsupervised Building Damage Detection by Roger Garay Avendaño is a document available to read on EtoBox.
The document discusses a novel unsupervised classification method called IHF to detect collapsed buildings from satellite imagery after disasters without requiring training samples. IHF uses hazard distribution data and fragility functions to calculate damage probabilities for buildings, which are then used as inputs for classification instead of training samples. The method is applied to the 2011 Tohoku earthquake and tsunami, detecting collapsed buildings with over 80% accuracy compared to field surveys.
- Author
- Roger Garay Avendaño
- Language
- EN