| spelling |
Erickson Rangel do Nascimentohttp://lattes.cnpq.br/6900352659470721Mário Fernando Montenegro Camposhttp://lattes.cnpq.br/5792248901353552Paulo Lilles Jorge Drews JuniorFlávio Luis Cardeal Páduahttp://lattes.cnpq.br/2761874662555561Walysson Vital Barbosa2022-08-10T19:13:30Z2022-08-10T19:13:30Z2020-02-04http://hdl.handle.net/1843/44155Advances in technology have allowed humans to delve into the depths of Earth and to study the outer space, even if our resources are not sufficient to help us answer all questions about each one of these environments. The underwater environment is one of those places, which has been vastly studied in past years due to the increasing use of underwater research locations. However, there are a few reasons why studying this environment is challenging. In most cases, structures located underwater cannot be moved out of this medium as they can lose their properties and be damaged. Moreover, images taken in these environments have very poor quality in comparison to images from out of water places. The water medium causes various effects during the image acquisition process. Rays of light are scattered and absorbed as they travel to the camera. This thesis proposes an underwater image restoration method based on convolutional neural networks and image quality metrics, the former being considered universal function approximators. Features extracted from the original underwater image are applied to the inverse image formation model in order to recover the original image radiance. No labeled data is needed as the network is trained based only in the quality metrics computed using the original and restored underwater images. In 60% of the cases, our proposed methodology performs better than the techniques applied to the improvement of underwater images, taking into consideration the UCIQE metric. Additionally, two underwater image datasets are presented, which were acquired on a planned process, focusing on underwater image restoration purposes.O avanço tecnológico tem nos permitido extrair informações e analisar os mais variados tipos de ambientes. O meio subaquático está incluído nesse conjunto de lugares e tem sido amplamente estudado nos últimos anos devido a áreas emergentes de pesquisas subaquáticas. No entanto, existem algumas razões pelas quais estudar neste ambiente se torna um desafio. Estruturas presentes debaixo d’água, como as de sítios arqueológicos, muitas vezes não podem ser movidas para fora desse meio, pois podem perder suas propriedades e, consequentemente, serem danificadas. Além disso, imagens tiradas nesses ambientes possuem qualidade muito baixa em comparação com imagens de fora d’água. O ambiente subaquático causa diversos efeitos durante o processo de aquisição da imagem. Raios de luz são espalhados e absorvidos enquanto viajam até o sensor da câmera. A presente dissertação propõe um método de restauração de imagens de cenas subaquáticas baseado na extração de parâmetros utilizando redes neurais convolucionais (CNNs) combinada com métricas de qualidade de imagem. Os parâmetros extraídos da imagem subaquática original são aplicados ao modelo de formação da imagem para recuperar a radiância original da imagem. Não são necessários dados rotulados, já que a rede é treinada com base apenas nas métricas de qualidade calculadas usando as imagens subaquáticas original e restaurada. A metodologia proposta se sobressaiu em 60% dos casos em comparação às demais abordagens apresentadas quando aplicadas na restauração de imagens subaquáticas, levando em consideração a métrica UCIQE. Além disso, dois conjuntos de imagens subaquáticas são apresentados, adquiridos num processo planejado e direcionado ao problema de restauração de imagens subaquáticas.CNPq - Conselho Nacional de Desenvolvimento Científico e TecnológicoFAPEMIG - Fundação de Amparo à Pesquisa do Estado de Minas GeraisCAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível SuperiorengUniversidade Federal de Minas GeraisPrograma de Pós-Graduação em Ciência da ComputaçãoUFMGBrasilICX - DEPARTAMENTO DE CIÊNCIA DA COMPUTAÇÃOComputação – TesesProcessamento de imagens -- Tecnicas digitas – Restauração e conservação -TesesVisão subaquática – TesesRedes neurais convolucionais – Tesesimage restorationunderwater visionconvolutional neural networksimage quality metricsQuality-driven end-to-end restoration system for underwater images using a self-supervised convolutional neural networkSistema para restauração de imagens subaquáticas ponta a ponta orientado pela qualidade usando uma rede neural convolucional auto-supervisionadainfo:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/masterThesisinfo:eu-repo/semantics/openAccessreponame:Repositório Institucional da UFMGinstname:Universidade Federal de Minas Gerais (UFMG)instacron:UFMGORIGINAL2020_barbosa_thesis_final.pdf2020_barbosa_thesis_final.pdfapplication/pdf2533875https://repositorio.ufmg.br/bitstream/1843/44155/1/2020_barbosa_thesis_final.pdf0c40b7ca2dadeef8590b035233d322a5MD51LICENSElicense.txtlicense.txttext/plain; charset=utf-82118https://repositorio.ufmg.br/bitstream/1843/44155/2/license.txtcda590c95a0b51b4d15f60c9642ca272MD521843/441552022-08-10 16:13:31.086oai:repositorio.ufmg.br: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ório InstitucionalPUBhttps://repositorio.ufmg.br/oaiopendoar:2022-08-10T19:13:31Repositório Institucional da UFMG - Universidade Federal de Minas Gerais (UFMG)false
|