YOLO Convolutional neural network for building damage detection in hydrometeorological disasters using satellite imagery
Natural disasters pose a continuous threat to populations worldwide, with hydrometeorological disasters standing out due to their unpredictability, rapid onset, and significant destructive capacity. Consequently, countries invest substantial resources in implementing pre- and post-disaster measures,...
- Autores:
-
Moreno González, César Luis
- Tipo de recurso:
- Trabajo de grado de pregrado
- Fecha de publicación:
- 2024
- Institución:
- Universidad de los Andes
- Repositorio:
- Séneca: repositorio Uniandes
- Idioma:
- eng
- OAI Identifier:
- oai:repositorio.uniandes.edu.co:1992/74680
- Acceso en línea:
- https://hdl.handle.net/1992/74680
- Palabra clave:
- Machine Learning
Deep Learning
Computer Vision
Detection Models
Natural Disasters
Hydrometeorological Disasters
Ingeniería
- Rights
- embargoedAccess
- License
- https://repositorio.uniandes.edu.co/static/pdf/aceptacion_uso_es.pdf
| id |
UNIANDES2_a6fdfbdefe3fe9cdfabdb0aabc9c3d43 |
|---|---|
| oai_identifier_str |
oai:repositorio.uniandes.edu.co:1992/74680 |
| network_acronym_str |
UNIANDES2 |
| network_name_str |
Séneca: repositorio Uniandes |
| repository_id_str |
|
| dc.title.none.fl_str_mv |
YOLO Convolutional neural network for building damage detection in hydrometeorological disasters using satellite imagery |
| title |
YOLO Convolutional neural network for building damage detection in hydrometeorological disasters using satellite imagery |
| spellingShingle |
YOLO Convolutional neural network for building damage detection in hydrometeorological disasters using satellite imagery Machine Learning Deep Learning Computer Vision Detection Models Natural Disasters Hydrometeorological Disasters Ingeniería |
| title_short |
YOLO Convolutional neural network for building damage detection in hydrometeorological disasters using satellite imagery |
| title_full |
YOLO Convolutional neural network for building damage detection in hydrometeorological disasters using satellite imagery |
| title_fullStr |
YOLO Convolutional neural network for building damage detection in hydrometeorological disasters using satellite imagery |
| title_full_unstemmed |
YOLO Convolutional neural network for building damage detection in hydrometeorological disasters using satellite imagery |
| title_sort |
YOLO Convolutional neural network for building damage detection in hydrometeorological disasters using satellite imagery |
| dc.creator.fl_str_mv |
Moreno González, César Luis |
| dc.contributor.advisor.none.fl_str_mv |
Lozano Garzón, Carlos Andrés Montoya Orozco, Germán Adolfo |
| dc.contributor.author.none.fl_str_mv |
Moreno González, César Luis |
| dc.subject.keyword.eng.fl_str_mv |
Machine Learning Deep Learning Computer Vision Detection Models Natural Disasters Hydrometeorological Disasters |
| topic |
Machine Learning Deep Learning Computer Vision Detection Models Natural Disasters Hydrometeorological Disasters Ingeniería |
| dc.subject.themes.spa.fl_str_mv |
Ingeniería |
| description |
Natural disasters pose a continuous threat to populations worldwide, with hydrometeorological disasters standing out due to their unpredictability, rapid onset, and significant destructive capacity. Consequently, countries invest substantial resources in implementing pre- and post-disaster measures, including prevention, and reconstruction. Developing countries, however, face stronger budgetary constraints and continuously depend on international support, having a limited and fluctuating ability to implement optimal disaster response strategies, affecting their response capabilities and timeliness, both of which are crucial for saving lives. This paper addressed these challenges by training and implementing YOLO (You Only Look Once) Convolutional Neural Network version 8 models, using high-resolution satellite images from the Maxar GeoEye-1 satellite. Additionally, this paper tested the viability of publicly publishing the model to be accessed via a REST API, providing a low-cost tool for quickly identifying damaged buildings following natural disasters, and enhancing the efficiency and effectiveness of disaster response efforts. |
| publishDate |
2024 |
| dc.date.accessioned.none.fl_str_mv |
2024-07-24T19:09:53Z |
| dc.date.issued.none.fl_str_mv |
2024-07-12 |
| dc.date.accepted.none.fl_str_mv |
2024-07-24 |
| dc.type.none.fl_str_mv |
Trabajo de grado - Pregrado |
| dc.type.driver.none.fl_str_mv |
info:eu-repo/semantics/bachelorThesis |
| dc.type.version.none.fl_str_mv |
info:eu-repo/semantics/acceptedVersion |
| dc.type.coar.none.fl_str_mv |
http://purl.org/coar/resource_type/c_7a1f |
| dc.type.content.none.fl_str_mv |
Text |
| dc.type.redcol.none.fl_str_mv |
http://purl.org/redcol/resource_type/TP |
| format |
http://purl.org/coar/resource_type/c_7a1f |
| status_str |
acceptedVersion |
| dc.identifier.uri.none.fl_str_mv |
https://hdl.handle.net/1992/74680 |
| dc.identifier.instname.none.fl_str_mv |
instname:Universidad de los Andes |
| dc.identifier.reponame.none.fl_str_mv |
reponame:Repositorio Institucional Séneca |
| dc.identifier.repourl.none.fl_str_mv |
repourl:https://repositorio.uniandes.edu.co/ |
| url |
https://hdl.handle.net/1992/74680 |
| identifier_str_mv |
instname:Universidad de los Andes reponame:Repositorio Institucional Séneca repourl:https://repositorio.uniandes.edu.co/ |
| dc.language.iso.none.fl_str_mv |
eng |
| language |
eng |
| dc.relation.references.none.fl_str_mv |
Azure Machine Learning Pricing, https://azure.microsoft.com/en-us/pricing/details/machine-learning/, last accessed 2024/07/04 Bermúdez, L.Learning from Hurricane Maria’s Impacts on Puerto Rico. In National Institute of Standards and Technology Special Publication. NIST SP 1262. (2021) Calton, Landon & Wei, Zhangping. (2022). Using Artificial Neural Network Models to Assess Hurricane Damage through Transfer Learning. Applied Sciences. 12. 1466. https://doi.org/10.3390/app12031466 COCO - Common objects in context, https://cocodataset.org/#home, last accessed 2024/03/25 Disaster risk management, https://www.bancomundial.org/es/topic/disasterriskmanagement/overview#:~:text=Desde%201980%2C%20a%20nivel%20mundial,cercanas%20a%20USD%206%20billones, last accessed 2024/06/26 IBM Analytics Solution Unified Method, http://gforge.icesi.edu.co/ASUM-DM_External/index.htm#cognos.external.asum-DM_Teaser/deliveryprocesses/M-DM_8A5C87D5.html_desc.html?proc=_0eKIHlt6EeW_y7k3h2HTng&path=_0eKIHlt6EeW_y7k3h2HTng, last accessed 2024/06/26 IBM. Foundational Methodology for Data Science. (2015) May, S. - Dupuis, A. - Lagrange, A. - De Vieilleville, F. - Fernandez, Martin, C:Building damage assessment with deep learning, 1133–1138. France (2022) GeoEye-1, https://resources.maxar.com/data-sheets/geoeye-1, last accessed 2024/05/05 NASA Joins Forces with Developing Nations to Reduce Disaster Risk, https://appliedsciences.nasa.gov/our-impact/news/nasa-joins-forces-developing-nations-reduce-disaster-risk, last accessed 2024/05/15 Rao, A., Jung, J., Silva, V., Molinario, G., and Yun, S.-H.: Earthquake building damage detection based on synthetic-aperture-radar imagery and machine learning, Nat. Hazards Earth Syst. Sci., 23, 789–807, (2023). https://doi.org/10.5194/nhess-23-789-2023 Redmon, J.- Divvala, S. - Girshick, R. - Farhadi, A: You only look once: Unified,real-time object detection. In Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 779–788. Las Vegas (2016) Srivastava, S., Divekar, A.V., Anilkumar, C. et al. Comparative analysis of deep learning image detection algorithms. J Big Data 8, 66 (2021). https://doi.org/10.1186/s40537-021-00434-w Transparency Portal. https://recovery.pr.gov/en/huracanes, last accessed 2024/06/24 United States Agency for International Development. Hydrometeorological Hazards Sector Update. (2019) United Nations Climate Change Secretariat. How developing countries are addressing hazards, focusing on relevant lessons learned and good practices. (2020) United Nations Frame Convention on Climate Change - UNFCCC Fourth synthesis of technology needs identified by Parties not included in Annex I to the Convention. (2020) United Nations International Strategy for Disaster Reduction (2019) Flash Flood. https://www.undrr.org/understanding-disaster-risk/terminology/hips/mh0006, last accessed 2024/06/24 Ultralytics YOLO Docs - Object Detection https://docs.ultralytics.com/tasks/detect/, last accessed 2024/06/27 Ultralytics YOLO Docs - Performance Metrics Deep Dive https://docs.ultralytics.com/guides/yolo-performance-metrics/#class-wise-metrics, last accessed 2024/06/20 Ultralytics YOLO Docs - Tips for Best Training Results, https://docs.ultralytics.com/yolov5/tutorials/tips_for_best_training_results/, last accessed 2024/06/22 xView2: Assess Building Damage, https://xview2.org/dataset, last accessed 2024/05/25 2024 EY Open Science Data Challenge: Coastal Resilience https://challenge.ey.com/2024, last accessed 2024/04/04 |
| dc.rights.uri.none.fl_str_mv |
https://repositorio.uniandes.edu.co/static/pdf/aceptacion_uso_es.pdf |
| dc.rights.accessrights.none.fl_str_mv |
info:eu-repo/semantics/embargoedAccess |
| dc.rights.coar.none.fl_str_mv |
http://purl.org/coar/access_right/c_f1cf |
| rights_invalid_str_mv |
https://repositorio.uniandes.edu.co/static/pdf/aceptacion_uso_es.pdf http://purl.org/coar/access_right/c_f1cf |
| eu_rights_str_mv |
embargoedAccess |
| dc.format.extent.none.fl_str_mv |
16 páginas |
| dc.format.mimetype.none.fl_str_mv |
application/pdf |
| dc.publisher.none.fl_str_mv |
Universidad de los Andes |
| dc.publisher.program.none.fl_str_mv |
Ingeniería de Sistemas y Computación |
| dc.publisher.faculty.none.fl_str_mv |
Facultad de Ingeniería |
| dc.publisher.department.none.fl_str_mv |
Departamento de Ingeniería de Sistemas y Computación |
| publisher.none.fl_str_mv |
Universidad de los Andes |
| institution |
Universidad de los Andes |
| bitstream.url.fl_str_mv |
https://repositorio.uniandes.edu.co/bitstreams/a6a3ce2d-1ede-44a3-9259-2884b791d434/download https://repositorio.uniandes.edu.co/bitstreams/edfc4c6b-372b-452d-912e-5bbb96e53583/download https://repositorio.uniandes.edu.co/bitstreams/1d2d41cb-3c80-4f02-90c7-1e30560a52d7/download |
| bitstream.checksum.fl_str_mv |
ae9e573a68e7f92501b6913cc846c39f 5e67690dbe380d6f5c696fda05e66ef5 2b876a959c1ef41f91a843db21077f82 |
| bitstream.checksumAlgorithm.fl_str_mv |
MD5 MD5 MD5 |
| repository.name.fl_str_mv |
Repositorio institucional Séneca |
| repository.mail.fl_str_mv |
adminrepositorio@uniandes.edu.co |
| _version_ |
1808390464490962944 |
| spelling |
Lozano Garzón, Carlos Andrésvirtual::19121-1Montoya Orozco, Germán Adolfovirtual::19122-1Moreno González, César Luis2024-07-24T19:09:53Z2024-07-122024-07-24https://hdl.handle.net/1992/74680instname:Universidad de los Andesreponame:Repositorio Institucional Sénecarepourl:https://repositorio.uniandes.edu.co/Natural disasters pose a continuous threat to populations worldwide, with hydrometeorological disasters standing out due to their unpredictability, rapid onset, and significant destructive capacity. Consequently, countries invest substantial resources in implementing pre- and post-disaster measures, including prevention, and reconstruction. Developing countries, however, face stronger budgetary constraints and continuously depend on international support, having a limited and fluctuating ability to implement optimal disaster response strategies, affecting their response capabilities and timeliness, both of which are crucial for saving lives. This paper addressed these challenges by training and implementing YOLO (You Only Look Once) Convolutional Neural Network version 8 models, using high-resolution satellite images from the Maxar GeoEye-1 satellite. Additionally, this paper tested the viability of publicly publishing the model to be accessed via a REST API, providing a low-cost tool for quickly identifying damaged buildings following natural disasters, and enhancing the efficiency and effectiveness of disaster response efforts.Pregrado16 páginasapplication/pdfengUniversidad de los AndesIngeniería de Sistemas y ComputaciónFacultad de IngenieríaDepartamento de Ingeniería de Sistemas y Computaciónhttps://repositorio.uniandes.edu.co/static/pdf/aceptacion_uso_es.pdfinfo:eu-repo/semantics/embargoedAccesshttp://purl.org/coar/access_right/c_f1cfYOLO Convolutional neural network for building damage detection in hydrometeorological disasters using satellite imageryTrabajo de grado - Pregradoinfo:eu-repo/semantics/bachelorThesisinfo:eu-repo/semantics/acceptedVersionhttp://purl.org/coar/resource_type/c_7a1fTexthttp://purl.org/redcol/resource_type/TPMachine LearningDeep LearningComputer VisionDetection ModelsNatural DisastersHydrometeorological DisastersIngenieríaAzure Machine Learning Pricing, https://azure.microsoft.com/en-us/pricing/details/machine-learning/, last accessed 2024/07/04Bermúdez, L.Learning from Hurricane Maria’s Impacts on Puerto Rico. In National Institute of Standards and Technology Special Publication. NIST SP 1262. (2021)Calton, Landon & Wei, Zhangping. (2022). Using Artificial Neural Network Models to Assess Hurricane Damage through Transfer Learning. Applied Sciences. 12. 1466. https://doi.org/10.3390/app12031466COCO - Common objects in context, https://cocodataset.org/#home, last accessed 2024/03/25Disaster risk management, https://www.bancomundial.org/es/topic/disasterriskmanagement/overview#:~:text=Desde%201980%2C%20a%20nivel%20mundial,cercanas%20a%20USD%206%20billones, last accessed 2024/06/26IBM Analytics Solution Unified Method, http://gforge.icesi.edu.co/ASUM-DM_External/index.htm#cognos.external.asum-DM_Teaser/deliveryprocesses/M-DM_8A5C87D5.html_desc.html?proc=_0eKIHlt6EeW_y7k3h2HTng&path=_0eKIHlt6EeW_y7k3h2HTng, last accessed 2024/06/26IBM. Foundational Methodology for Data Science. (2015)May, S. - Dupuis, A. - Lagrange, A. - De Vieilleville, F. - Fernandez, Martin, C:Building damage assessment with deep learning, 1133–1138. France (2022)GeoEye-1, https://resources.maxar.com/data-sheets/geoeye-1, last accessed 2024/05/05NASA Joins Forces with Developing Nations to Reduce Disaster Risk, https://appliedsciences.nasa.gov/our-impact/news/nasa-joins-forces-developing-nations-reduce-disaster-risk, last accessed 2024/05/15Rao, A., Jung, J., Silva, V., Molinario, G., and Yun, S.-H.: Earthquake building damage detection based on synthetic-aperture-radar imagery and machine learning,Nat. Hazards Earth Syst. Sci., 23, 789–807, (2023). https://doi.org/10.5194/nhess-23-789-2023Redmon, J.- Divvala, S. - Girshick, R. - Farhadi, A: You only look once: Unified,real-time object detection. In Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 779–788. Las Vegas (2016)Srivastava, S., Divekar, A.V., Anilkumar, C. et al. Comparative analysis of deep learning image detection algorithms. J Big Data 8, 66 (2021). https://doi.org/10.1186/s40537-021-00434-wTransparency Portal. https://recovery.pr.gov/en/huracanes, last accessed 2024/06/24United States Agency for International Development. Hydrometeorological Hazards Sector Update. (2019)United Nations Climate Change Secretariat. How developing countries are addressing hazards, focusing on relevant lessons learned and good practices. (2020)United Nations Frame Convention on Climate Change - UNFCCC Fourth synthesis of technology needs identified by Parties not included in Annex I to the Convention. (2020)United Nations International Strategy for Disaster Reduction (2019) Flash Flood. https://www.undrr.org/understanding-disaster-risk/terminology/hips/mh0006, last accessed 2024/06/24Ultralytics YOLO Docs - Object Detection https://docs.ultralytics.com/tasks/detect/, last accessed 2024/06/27Ultralytics YOLO Docs - Performance Metrics Deep Dive https://docs.ultralytics.com/guides/yolo-performance-metrics/#class-wise-metrics, last accessed 2024/06/20Ultralytics YOLO Docs - Tips for Best Training Results, https://docs.ultralytics.com/yolov5/tutorials/tips_for_best_training_results/, last accessed 2024/06/22xView2: Assess Building Damage, https://xview2.org/dataset, last accessed 2024/05/252024 EY Open Science Data Challenge: Coastal Resilience https://challenge.ey.com/2024, last accessed 2024/04/04201924899Publication9a0ca46c-ed4d-4da2-af46-db6aa9454a0dvirtual::19121-1a197a9f7-96e5-47cb-a497-2ee4c9cdce71virtual::19122-19a0ca46c-ed4d-4da2-af46-db6aa9454a0dvirtual::19121-1a197a9f7-96e5-47cb-a497-2ee4c9cdce71virtual::19122-1https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh=0000219541virtual::19121-1LICENSElicense.txtlicense.txttext/plain; charset=utf-82535https://repositorio.uniandes.edu.co/bitstreams/a6a3ce2d-1ede-44a3-9259-2884b791d434/downloadae9e573a68e7f92501b6913cc846c39fMD51ORIGINALDevelopment_of_Machine_Learning_models_for_the_identification_of_buildings_affected_by_hydrometeorological_disasters (4).pdfDevelopment_of_Machine_Learning_models_for_the_identification_of_buildings_affected_by_hydrometeorological_disasters (4).pdfEl trabajo de grado debe ser privado de forma indefinida, pues está postulado para ser publicado en la conferencia ICAI 2024 7th International Conference on Applied Informatics. La publicación pública en este repositorio puede descartarlo para su participación en la conferencia.application/pdf429118https://repositorio.uniandes.edu.co/bitstreams/edfc4c6b-372b-452d-912e-5bbb96e53583/download5e67690dbe380d6f5c696fda05e66ef5MD52autorizacion tesis Cesar Moreno.pdfautorizacion tesis Cesar Moreno.pdfHIDEapplication/pdf279408https://repositorio.uniandes.edu.co/bitstreams/1d2d41cb-3c80-4f02-90c7-1e30560a52d7/download2b876a959c1ef41f91a843db21077f82MD531992/74680oai:repositorio.uniandes.edu.co:1992/746802024-07-24 14:14:04.142https://repositorio.uniandes.edu.co/static/pdf/aceptacion_uso_es.pdfrestrictedhttps://repositorio.uniandes.edu.coRepositorio institucional Sénecaadminrepositorio@uniandes.edu.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 |