Extraction of Earthquake Damage Information and Mapping of Buildings from Single Post-earthquake Polarimetric Synthetic Aperture Radar Image Based on Polarimetric Decomposition and Texture Features

Wei Zhai, Xiaoqing Wang, Y Bi, Jun Liu, Guiyu Zhu, Jianqing Du

Research output: Contribution to journalArticlepeer-review

12 Downloads (Pure)


The collapse of buildings caused by destructive earthquakes often leads to severe casualties and economic losses. After an earthquake, an accurate assessment of building damage will be essential in making plans of emergency responses. Four-polarimetric synthetic aperture radar (PolSAR) data have advantages over synthetic aperture radar (SAR) imaging data, because they are not occluded by sunlight or clouds. They also contain the most abundant information of four polarimetric channels. Therefore, a single PolSAR image can be used to identify post-earthquake building damage. It is easy to overestimate the number of collapsed buildings and the degree of damage by earthquakes when using only a traditional polarimetric decomposition method for PolSAR data. In urban areas, buildings can stand in parallel in typical SAR imaging with strong scattering features, and there are also some oriented standing buildings with lower scattering intensity or similar scattering characteristics to collapsed buildings; thus, these oriented standing buildings are often misconstrued as collapsed buildings. In this study, we propose a new texture feature, namely, the mean standard deviation (MSD) index based on the gray-level co-occurrence matrix (GLCM), to solve the overestimation of building damage caused by earthquakes. Moreover, on the basis of the improved Yamaguchi four-component decomposition method and the MSD index, we develop a method of identifying the damage of buildings using only a single post-earthquake PolSAR image. In our study case, 75000 undamaged and damaged building samples are used in the experiment. The proposed method has greatly improved the accuracy and reliability of extracted building damage information. The experimental results show identification accuracies of 82.43 and 80.30% for damaged and undamaged buildings, respectively. Compared with the traditional polarimetric decomposition method, 66.89% standing buildings are successfully isolated from the mixture of collapsed buildings using our method.
Original languageEnglish
Article number12
Pages (from-to)4451-4462
Number of pages12
JournalSensors and Materials
Issue number12
Publication statusPublished (in print/issue) - 15 Dec 2022

Bibliographical note

Funding Information:
This work was supported by the National key R&D program of China (No. 2017YFB0504104); the Gansu Province Science and Technology Program (22JR5RA822); the National Natural Science Foundation of China (41601479; 42061073); the Key Talent Project of Gansu Province (11276679015); the Dragon 5 programme (ID: 59308), a collaboration between the European Space Agency and the Ministry of Science and Technology of China; the High Score Project of State Administration of Science, Technology and Industry for National Defense (Phase II): Application of remote sensing based seismic intensity and loss assessment (31-Y30F09-9001-20/22-12); the Science for Earthquake Resilience of China Earthquake Administration (XH18049); the Basic Research Project of Institute of Earthquake Science, China Earthquake Administration (2021IESLZ4); the Gansu Earthquake Administration innovation team special fund (2019TD-01-02); the State Scholarship Fund of China Scholarship Council (CSC); and the Earthquake Science and Technology Development Fund Program of Lan-zhou Earthquake Research Institute, China Earthquake Administration (2015M02).

Publisher Copyright:
© MYU K.K.


  • buildings
  • earthquake damage assessment
  • polarimetric decomposition
  • PoISAR
  • texture features
  • PolSAR


Dive into the research topics of 'Extraction of Earthquake Damage Information and Mapping of Buildings from Single Post-earthquake Polarimetric Synthetic Aperture Radar Image Based on Polarimetric Decomposition and Texture Features'. Together they form a unique fingerprint.

Cite this