Please use this identifier to cite or link to this item: http://hdl.handle.net/10397/8112
Title: A method to detect earthquake-collapsed buildings from high-resolution satellite images
Authors: Shi, W 
Hao, M
Issue Date: 2013
Source: Remote sensing letters, 2013, v. 4, no. 12, p. 1166-1175 How to cite?
Journal: Remote Sensing Letters 
Abstract: A method is proposed to detect collapsed buildings induced by earthquakes. The method computes quantitative indices of the spectral responses for the city plan objects (buildings), assuming that a building was collapsed if the spectral response is sufficiently heterogeneous. First, the pre-earthquake boundaries of buildings stored in the geographic information system (GIS) vector data are used as a reference for determining the extent of each building in the post-image. Second, an improved active contour model is implemented to extract those homogeneous regions in the building boundaries on the post-earthquake image. Third, the shape similarity index (SSI) between the extracted homogeneous region and the corresponding pre-building boundary is calculated, and the area ratio index (ARI) between the extracted homogeneous pixel areas and the true reference pixel areas is calculated. Finally, the k-means clustering method is implemented to partition buildings into collapsed and undamaged sections based on the SSI and ARI. The experimental results indicate the strong robustness and high effectiveness of the proposed method. It is worth noting that a threshold is not needed to determine whether buildings are collapsed. The method quickly and accurately provides information on collapsed buildings.
URI: http://hdl.handle.net/10397/8112
ISSN: 2150-704X
DOI: 10.1080/2150704X.2013.858839
Appears in Collections:Journal/Magazine Article

Access
View full-text via PolyU eLinks SFX Query
Show full item record

SCOPUSTM   
Citations

8
Citations as of Jan 23, 2017

WEB OF SCIENCETM
Citations

8
Last Week
1
Last month
1
Citations as of Jan 20, 2017

Page view(s)

16
Last Week
0
Last month
Checked on Jan 22, 2017

Google ScholarTM

Check

Altmetric



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.