A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
UAV-Based Post-disaster Assessment of Residential Buildings Using Image Processing
https://orcid.org/http://orcid.org/0000-0002-8387-2036
https://orcid.org/http://orcid.org/0000-0003-3322-5583
https://orcid.org/http://orcid.org/0000-0002-7750-0906
This paper puts forward an image-processing-based method for detecting damaged buildings using post-disaster imagery data. Depending on the level of a disaster, manual inspections could be time-consuming, expensive, and risky. Rapid assessment of damaged buildings is crucial for proper rescue and reconstruction planning in the aftermath of a disaster. As a common image collection tool after disasters, this study focuses on using unmanned aerial vehicles (UAVs), ensuring the capability of the proposed method for automatic assessments. Most of the previous studies on rapid building damage detections required pre-disaster images as inputs, which limits their application to locations with such prior information. However, this study utilized only the post-disaster images to develop a tool capable of detecting damage level of buildings. Insufficient imagery data was the main challenge facing previous studies in this field. Thus, novel features are defined based on edge detection and texture analysis of images to capture damages in the buildings more accurately and decreasing the need for larger training datasets. These features are then employed in a Naïve Bayesian classification method to account for the prevailing uncertainties. The proposed method is verified using real-life images. Such model, in conjunction with an automatic data collection using UAVs, will make it possible to have a rapid assessment and prioritization for rescue and reconstruction planning in the aftermath of a disaster.
UAV-Based Post-disaster Assessment of Residential Buildings Using Image Processing
https://orcid.org/http://orcid.org/0000-0002-8387-2036
https://orcid.org/http://orcid.org/0000-0003-3322-5583
https://orcid.org/http://orcid.org/0000-0002-7750-0906
This paper puts forward an image-processing-based method for detecting damaged buildings using post-disaster imagery data. Depending on the level of a disaster, manual inspections could be time-consuming, expensive, and risky. Rapid assessment of damaged buildings is crucial for proper rescue and reconstruction planning in the aftermath of a disaster. As a common image collection tool after disasters, this study focuses on using unmanned aerial vehicles (UAVs), ensuring the capability of the proposed method for automatic assessments. Most of the previous studies on rapid building damage detections required pre-disaster images as inputs, which limits their application to locations with such prior information. However, this study utilized only the post-disaster images to develop a tool capable of detecting damage level of buildings. Insufficient imagery data was the main challenge facing previous studies in this field. Thus, novel features are defined based on edge detection and texture analysis of images to capture damages in the buildings more accurately and decreasing the need for larger training datasets. These features are then employed in a Naïve Bayesian classification method to account for the prevailing uncertainties. The proposed method is verified using real-life images. Such model, in conjunction with an automatic data collection using UAVs, will make it possible to have a rapid assessment and prioritization for rescue and reconstruction planning in the aftermath of a disaster.
UAV-Based Post-disaster Assessment of Residential Buildings Using Image Processing
https://orcid.org/http://orcid.org/0000-0002-8387-2036
https://orcid.org/http://orcid.org/0000-0003-3322-5583
https://orcid.org/http://orcid.org/0000-0002-7750-0906
This paper puts forward an image-processing-based method for detecting damaged buildings using post-disaster imagery data. Depending on the level of a disaster, manual inspections could be time-consuming, expensive, and risky. Rapid assessment of damaged buildings is crucial for proper rescue and reconstruction planning in the aftermath of a disaster. As a common image collection tool after disasters, this study focuses on using unmanned aerial vehicles (UAVs), ensuring the capability of the proposed method for automatic assessments. Most of the previous studies on rapid building damage detections required pre-disaster images as inputs, which limits their application to locations with such prior information. However, this study utilized only the post-disaster images to develop a tool capable of detecting damage level of buildings. Insufficient imagery data was the main challenge facing previous studies in this field. Thus, novel features are defined based on edge detection and texture analysis of images to capture damages in the buildings more accurately and decreasing the need for larger training datasets. These features are then employed in a Naïve Bayesian classification method to account for the prevailing uncertainties. The proposed method is verified using real-life images. Such model, in conjunction with an automatic data collection using UAVs, will make it possible to have a rapid assessment and prioritization for rescue and reconstruction planning in the aftermath of a disaster.
UAV-Based Post-disaster Assessment of Residential Buildings Using Image Processing
Lecture Notes in Civil Engineering
Desjardins, Serge (editor) / Poitras, Gérard J. (editor) / El Damatty, Ashraf (editor) / Elshaer, Ahmed (editor) / Shirzad-Ghaleroudkhani, Nima (author) / Jozi, Daniel (author) / Luhadia, Garvit (author) / Abtahi, Shaghayegh (author) / Gül, Mustafa (author)
Canadian Society of Civil Engineering Annual Conference ; 2023 ; Moncton, NB, Canada
2024-09-03
13 pages
Article/Chapter (Book)
Electronic Resource
English
Multi-Objective Framework for Resilient Post-Disaster Self-Built Residential Buildings
| DOAJ | 2024
Machine learning application to disaster damage repair cost modelling of residential buildings
| Taylor & Francis Verlag | 2025