- 103 views
- 99 downloads
UAV-based Post-disaster Damage Assessment of Buildings Using Image Processing
-
- Author / Creator
- Jozi, Seyed Danial
-
The extensive damages caused by natural disasters incur substantial costs to the built environment. The escalating frequency and severity of disasters, particularly hurricanes, driven by the impacts of Climate Change, highlight the urgency for prompt post-disaster assessments. A thorough and rapid post-disaster assessment plays a crucial role in facilitating the swift evaluation of the situation, and enables the determination of the extent of damage for each component in the built environment. To address this need, this study proposes an image-processing-based method utilizing Unmanned Aerial Vehicle (UAV) imagery of solely the post-disaster situation for the automated evaluation of the building damage. The suggested approach accordingly utilizes single post-disaster imagery of the buildings, and integrates texture-based features, encompassing texture dissimilarity and homogeneity, along with edge-based features. Canny edge detection is employed to introduce novel indices that gauge irregularity by assessing entropy of the detected edges and uniformity in the distribution of edge line orientations. These features are then input into a Naïve Bayesian Classification process, allowing for the classification of damaged and undamaged classes while accommodating the underlying uncertainties. The proposed method exhibits a validation accuracy of 91.3 percent when applied to unidentified building images, effectively distinguishing between damaged and undamaged structures. In addition, the functionality of the proposed method has been evaluated through application on a real-life post-disaster scene. The results underscore the potential efficacy of utilizing UAV-captured images and advanced image processing techniques for rapid and accurate post-disaster damage assessment.
-
- Subjects / Keywords
-
- Graduation date
- Fall 2024
-
- Type of Item
- Thesis
-
- Degree
- Master of Science
-
- License
- This thesis is made available by the University of Alberta Library with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.