Logo do repositório
 
Publicação

Application of artificial intelligence to the detection of foreign object debris at aerodromes’ movement area

2022Documento de Trabalho Acesso aberto
Ver registo simples
datacite.subject.fosEngenharia e Tecnologia
datacite.subject.fosEngenharia e Tecnologia::Engenharia Eletrotécnica, Eletrónica e Informática
datacite.subject.fosEngenharia e Tecnologia::Outras Engenharias e Tecnologias
datacite.subject.sdg09:Indústria, Inovação e Infraestruturas
datacite.subject.sdg11:Cidades e Comunidades Sustentáveis
datacite.subject.sdg03:Saúde de Qualidade
datacite.subject.sdg12:Produção e Consumo Sustentáveis
datacite.subject.sdg13:Ação Climática
datacite.subject.sdg17:Parcerias para a Implementação dos Objetivos
dc.contributor.authorAlmeida, João Miguel Brito de
dc.date.accessioned2026-05-05T18:47:18Z
dc.date.available2026-05-05T18:47:18Z
dc.date.issued2022
dc.descriptionResumo alargado da dissertação de mestrado com o mesmo título, defendida em 2002.
dc.description.abstractThe goal of the present dissertation is to develop a preliminary low-cost and passive system that detects Foreign Object Debris (FODs) at aerodromes based on computer vision with neural networks. FODs are a twofold problem, involving safety risks and high associated costs. Although some systems already exist to detect FODs, these are based on radars, making them expensive. We build a dataset of images to test the viability of this solution, which was already attempted by other authors but the datasets are not publicly available. Moreover, we build a simplified architecture of the system to capture the images. In parallel, we develop a software pipeline which starts with image capturing scripts and ends in the evaluation of the models of neural networks we selected. The datasets created result from three different electro-optical sensors: visible, near infrared and long-wave infrared. From the first, resulted a dataset of 9,260 images, from the second 5,672 and from the third 10,388. Our approach to this problem is based on supervised learning with image classification and object detection and we train the models in subsets of the datasets. For image classification, we choose Xception as the neural network, achieving an 98.86% accuracy. In the case of object detection, we opt for a single-stage detector – YOLOv3 –, achieving an AP of 91.08%. Finally, we test the same models on new examples and verify a decrease in their performance to 77.92% accuracy for the classifier and 37.49% AP for the detector.por
dc.identifier.citationAlmeida, J. M. B. (2022). Application of artificial intelligence to the detection of foreign object debris at aerodromes’ movement area. Unpublished manuscript.
dc.identifier.urihttp://hdl.handle.net/10400.26/62991
dc.language.isoeng
dc.peerreviewedn/a
dc.publisherAcademia da Força Aérea
dc.rights.urihttp://creativecommons.org/licenses/by-nc/4.0/
dc.subjectForeign object debris
dc.subjectComputer vision
dc.subjectDataset
dc.subjectImage classification
dc.subjectObject detection
dc.subjectDetritos de objetos estranhos
dc.subjectVisão por computador
dc.subjectConjunto de dado
dc.subjectClassificação de imagens
dc.subjectDeteção de objetos
dc.subjectAeródromos
dc.subjectSensores eletro-óticos
dc.subjectInteligência artificial
dc.subjectSegurança da aviação
dc.titleApplication of artificial intelligence to the detection of foreign object debris at aerodromes’ movement areaeng
dc.title.alternativeAplicação da inteligência artificial à deteção de detritos de objetos estranhos na área de movimento dos aeródromospor
dc.typeworking paper
dspace.entity.typePublication
oaire.versionhttp://purl.org/coar/version/c_be7fb7dd8ff6fe43

Ficheiros

Coleções

AFA - DA - ACPA - Documentos de Trabalho