Enter your email address
Submit
Write your message
Articles In Press
Back to the articles list |
Back to browse issues page
1- Shahid Chamran University of Ahvaz
Abstract: (3 Views)
Background and Aims: Natural disasters and military attacks can cause widespread destruction to buildings, making rapid and accurate damage assessment crucial for reducing casualties and speeding up rescue operations0. This study aims to present a novel method for assessing building damage using deep learning and image processing techniques.
Methodology: This research introduces a new method based on an encoder-decoder architecture that combines the DeepLabV3+ model with a self-attention mechanism. The xBD dataset, which contains pre- and post-disaster images from various incidents, was used to train and evaluate the model.
Findings: The proposed model successfully classified buildings into four damage levels: "no damage," "minor damage," "major damage," and "destroyed". It achieved superior results compared to previous methods with an overall F1-Score of 0.94.
Conclusion and Innovation: The primary innovation of this research lies in the simultaneous use of DeepLabV3+ capabilities, the attention mechanism, and data augmentation techniques, which significantly improved detection accuracy. The method can be applied in natural disaster management as well as in passive defense and military scenarios to quickly assess damage from attacks.
Methodology: This research introduces a new method based on an encoder-decoder architecture that combines the DeepLabV3+ model with a self-attention mechanism. The xBD dataset, which contains pre- and post-disaster images from various incidents, was used to train and evaluate the model.
Findings: The proposed model successfully classified buildings into four damage levels: "no damage," "minor damage," "major damage," and "destroyed". It achieved superior results compared to previous methods with an overall F1-Score of 0.94.
Conclusion and Innovation: The primary innovation of this research lies in the simultaneous use of DeepLabV3+ capabilities, the attention mechanism, and data augmentation techniques, which significantly improved detection accuracy. The method can be applied in natural disaster management as well as in passive defense and military scenarios to quickly assess damage from attacks.
Keywords: Building Damage, DeepLabV3+ Network, Multiclass Segmentation, UNet Network, Remote Sensing, Self-Attention Mechanism
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |
Copyright © 2025 The Author(s)" by "authors retain the copyright and full publishing rights"
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC-By-NC), which permits use, distribution, and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes.
Contact Information
