TY - JOUR
T1 - Object-based building damage assessment methodology using only post event ALOS-2/PALSAR-2 dual polarimetric SAR intensity images
AU - Bai, Yanbing
AU - Adriano, Bruno
AU - Mas, Erick
AU - Gokon, Hideomi
AU - Koshimura, Shunichi
N1 - Funding Information:
We would like to thank the Japan Aerospace Exploration Agency (JAXA) for providing our dataset of SAR imagery. This work was supported by JSPS KAKENHI Grant Number 25242035, the JST CREST Project, and the China Scholarship Council (CSC).
Publisher Copyright:
© 2017, Fuji Technology Press. All rights reserved.
PY - 2017/3
Y1 - 2017/3
N2 - Earthquake-induced building damage assessment is an indispensable prerequisite for disaster impact assessment, and the increasing availability of high-resolution Synthetic Aperture Radar (SAR) imagery has made it possible to construct damaged building inventories soon after earthquakes strike. However, the shortage of pre-seismic SAR datasets and the lack of available building footprint data pose challenges for rapid building damage assessment. Taking advantage of recent advances in machine learning algorithms, this study proposes an object-based building damage assessment methodology that uses only post-event SAR imagery. A Random Forest machine learning-based object classification, a simplified approach to the extraction of built-up areas, was developed and tested on two ALOS2/PALSAR-2 dual polarimetric SAR images acquired in affected areas soon after the 2015 Nepal earthquake. In addition, a series of texture metrics as well as the random scattering metric and reflection symmetry metric were found to significantly enhance classification accuracy. The feature selection was found to have a positive effect on overall performance. Moreover, the proposed Random Forest framework resulted in overall accuracies of 93% with a kappa coefficient of 0.885 when the object scale of 60 × 60 pixels and 15 features were adopted. A comparative experiment with the k-nearest neighbor framework demonstrated that the Random Forest framework is a significant step toward the achievement of a balanced, two-class classification.
AB - Earthquake-induced building damage assessment is an indispensable prerequisite for disaster impact assessment, and the increasing availability of high-resolution Synthetic Aperture Radar (SAR) imagery has made it possible to construct damaged building inventories soon after earthquakes strike. However, the shortage of pre-seismic SAR datasets and the lack of available building footprint data pose challenges for rapid building damage assessment. Taking advantage of recent advances in machine learning algorithms, this study proposes an object-based building damage assessment methodology that uses only post-event SAR imagery. A Random Forest machine learning-based object classification, a simplified approach to the extraction of built-up areas, was developed and tested on two ALOS2/PALSAR-2 dual polarimetric SAR images acquired in affected areas soon after the 2015 Nepal earthquake. In addition, a series of texture metrics as well as the random scattering metric and reflection symmetry metric were found to significantly enhance classification accuracy. The feature selection was found to have a positive effect on overall performance. Moreover, the proposed Random Forest framework resulted in overall accuracies of 93% with a kappa coefficient of 0.885 when the object scale of 60 × 60 pixels and 15 features were adopted. A comparative experiment with the k-nearest neighbor framework demonstrated that the Random Forest framework is a significant step toward the achievement of a balanced, two-class classification.
KW - 2015 nepal earthquake
KW - Object-based building damage assessment
KW - Post-event dual-polarimetric SAR imagery
KW - Random forest machine learning algorithms
UR - http://www.scopus.com/inward/record.url?scp=85015313967&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85015313967&partnerID=8YFLogxK
U2 - 10.20965/jdr.2017.p0259
DO - 10.20965/jdr.2017.p0259
M3 - Article
AN - SCOPUS:85015313967
VL - 12
SP - 259
EP - 271
JO - Journal of Disaster Research
JF - Journal of Disaster Research
SN - 1881-2473
IS - 2
ER -