A random forest model for seismic-damage buildings identification based on UAV images coupled with RFE and object-oriented methods

Haijia Wen (), Jiwei Hu, Fengguang Xiong, Chi Zhang, Chenhao Song and Xinzhi Zhou
Additional contact information
Haijia Wen: National Joint Engineering Research Center of Geohazards Prevention in the Reservoir Areas, Chongqing University
Jiwei Hu: National Joint Engineering Research Center of Geohazards Prevention in the Reservoir Areas, Chongqing University
Fengguang Xiong: National Joint Engineering Research Center of Geohazards Prevention in the Reservoir Areas, Chongqing University
Chi Zhang: National Joint Engineering Research Center of Geohazards Prevention in the Reservoir Areas, Chongqing University
Chenhao Song: National Joint Engineering Research Center of Geohazards Prevention in the Reservoir Areas, Chongqing University
Xinzhi Zhou: Tsing-Hua University

Natural Hazards: Journal of the International Society for the Prevention and Mitigation of Natural Hazards, 2023, vol. 119, issue 3, No 25, 1769 pages

Abstract: Abstract Earthquakes are sudden and unpredictable disasters that pose significant threats to human lives, property, and societal stability. Utilizing remote sensing techniques enables the rapid acquisition of disaster information from affected areas, aiding in the localization, identification, and serving as a reference for emergency response to damaged buildings. This study proposed a UAV image-based approach for seismic damage assessment in mountainous regions, combining recursive feature elimination (RFE) and object-oriented coupling with the random forest classifier. The ESP2 tool is employed to determine optimal segmentation scales, reducing the manual labor of segmenting images at various scales and improving the efficiency of selecting the best segmentation scale. Taking the June 18, 2019 post-earthquake UAV imagery of Shuanghe Town, Changning County, Sichuan Province as the study area, an object-oriented multi-level classification system is established using the ESP2 algorithm. Subsequently, features selected through recursive feature elimination (RFE) are utilized in conjunction with an optimized random forest classifier for building seismic damage identification. Results show that optimal segmentation scales for vegetation, rivers, and buildings are 196, 156, and 130, respectively, with RFE elimination process reducing the initial 57 features to 24. Following RFE optimization, the overall accuracy of building seismic damage identification reaches 95.3%, with a Kappa coefficient of 0.807. This approach, utilizing a reduced number of features for enhanced accuracy, effectively contributes to seismic damage identification for buildings.

Keywords: OBIA; Seismic damage; Buildings; UAV; RFE (search for similar items in EconPapers)
Date: 2023
References: Add references at CitEc
Citations:

Downloads: (external link)
http://link.springer.com/10.1007/s11069-023-06186-5 Abstract (text/html)
Access to the full text of the articles in this series is restricted.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:nathaz:v:119:y:2023:i:3:d:10.1007_s11069-023-06186-5

Ordering information: This journal article can be ordered from
http://www.springer.com/economics/journal/11069

DOI: 10.1007/s11069-023-06186-5

Access Statistics for this article

Natural Hazards: Journal of the International Society for the Prevention and Mitigation of Natural Hazards is currently edited by Thomas Glade, Tad S. Murty and Vladimír Schenk

More articles in Natural Hazards: Journal of the International Society for the Prevention and Mitigation of Natural Hazards from Springer, International Society for the Prevention and Mitigation of Natural Hazards
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().