Post-Disaster Building Damage Segmentation Using Convolutional Neural Networks

Natural disasters are events caused by nature such as earthquakes, tornadoes, tsunamis, forest fires, and others. The impacts of natural disasters are significant and varied across various sectors, including the economy, health, and primarily, infrastructure. Effective and efficient actions are nee...

Full description

Saved in:
Bibliographic Details
Main Authors: Revaldi Rahmatmulya, Agung Teguh Wibowo Almais, Mokhamad Amin Hariyadi
Format: Article
Language:English
Published: LPPM STIKI Malang 2025-07-01
Series:J-Intech (Journal of Information and Technology)
Subjects:
Online Access:https://jurnal.stiki.ac.id/J-INTECH/article/view/1919
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849426000908124160
author Revaldi Rahmatmulya
Agung Teguh Wibowo Almais
Mokhamad Amin Hariyadi
author_facet Revaldi Rahmatmulya
Agung Teguh Wibowo Almais
Mokhamad Amin Hariyadi
author_sort Revaldi Rahmatmulya
collection DOAJ
description Natural disasters are events caused by nature such as earthquakes, tornadoes, tsunamis, forest fires, and others. The impacts of natural disasters are significant and varied across various sectors, including the economy, health, and primarily, infrastructure. Effective and efficient actions are needed to assist in the recovery following natural disasters, one of which is aiding in the identification of building damage levels post-disaster. To address this issue, this research proposes a system capable of performing segmentation to determine the level of building damage post-natural disaster using convolutional neural network methods. The data utilized consists of aerial images sourced from xView2: Assess Building Damage, comprising 50 aerial images with 5 classes: no-damage, minor-damage, major-damage, destroyed, and unlabeled. The steps undertaken in this research include data preprocessing using patchify and data augmentation. Subsequently, feature extraction is performed using convolution, followed by the training process using a neural network with the proposed architecture. This study proposes an architecture with 27 hidden layers, with feature extraction utilizing average pooling. The model evaluation process will employ Mean Intersection over Union (MIoU) to assess how closely the segmentation prediction results resemble the original data. The proposed architecture demonstrates the best MIoU result with a value of 0.31 and an accuracy of 0.9577.
format Article
id doaj-art-1ce28fa3162246d68f66b44761476094
institution Kabale University
issn 2303-1425
2580-720X
language English
publishDate 2025-07-01
publisher LPPM STIKI Malang
record_format Article
series J-Intech (Journal of Information and Technology)
spelling doaj-art-1ce28fa3162246d68f66b447614760942025-08-20T03:29:35ZengLPPM STIKI MalangJ-Intech (Journal of Information and Technology)2303-14252580-720X2025-07-01130110.32664/j-intech.v13i01.1919Post-Disaster Building Damage Segmentation Using Convolutional Neural NetworksRevaldi Rahmatmulya0Agung Teguh Wibowo Almais1Mokhamad Amin Hariyadi2Universitas Islam Negeri Maulana Malik Ibrahim Malang, IndonesiaUniversitas Islam Negeri Maulana Malik Ibrahim Malang, IndonesiaUniversitas Islam Negeri Maulana Malik Ibrahim Malang, Indonesia Natural disasters are events caused by nature such as earthquakes, tornadoes, tsunamis, forest fires, and others. The impacts of natural disasters are significant and varied across various sectors, including the economy, health, and primarily, infrastructure. Effective and efficient actions are needed to assist in the recovery following natural disasters, one of which is aiding in the identification of building damage levels post-disaster. To address this issue, this research proposes a system capable of performing segmentation to determine the level of building damage post-natural disaster using convolutional neural network methods. The data utilized consists of aerial images sourced from xView2: Assess Building Damage, comprising 50 aerial images with 5 classes: no-damage, minor-damage, major-damage, destroyed, and unlabeled. The steps undertaken in this research include data preprocessing using patchify and data augmentation. Subsequently, feature extraction is performed using convolution, followed by the training process using a neural network with the proposed architecture. This study proposes an architecture with 27 hidden layers, with feature extraction utilizing average pooling. The model evaluation process will employ Mean Intersection over Union (MIoU) to assess how closely the segmentation prediction results resemble the original data. The proposed architecture demonstrates the best MIoU result with a value of 0.31 and an accuracy of 0.9577. https://jurnal.stiki.ac.id/J-INTECH/article/view/1919convolutional neural networkmachine learningsegmentasi
spellingShingle Revaldi Rahmatmulya
Agung Teguh Wibowo Almais
Mokhamad Amin Hariyadi
Post-Disaster Building Damage Segmentation Using Convolutional Neural Networks
J-Intech (Journal of Information and Technology)
convolutional neural network
machine learning
segmentasi
title Post-Disaster Building Damage Segmentation Using Convolutional Neural Networks
title_full Post-Disaster Building Damage Segmentation Using Convolutional Neural Networks
title_fullStr Post-Disaster Building Damage Segmentation Using Convolutional Neural Networks
title_full_unstemmed Post-Disaster Building Damage Segmentation Using Convolutional Neural Networks
title_short Post-Disaster Building Damage Segmentation Using Convolutional Neural Networks
title_sort post disaster building damage segmentation using convolutional neural networks
topic convolutional neural network
machine learning
segmentasi
url https://jurnal.stiki.ac.id/J-INTECH/article/view/1919
work_keys_str_mv AT revaldirahmatmulya postdisasterbuildingdamagesegmentationusingconvolutionalneuralnetworks
AT agungteguhwibowoalmais postdisasterbuildingdamagesegmentationusingconvolutionalneuralnetworks
AT mokhamadaminhariyadi postdisasterbuildingdamagesegmentationusingconvolutionalneuralnetworks