• Entrar
    Ver item 
    •   Página inicial
    • Teses & Dissertações
    • Teses & Dissertações
    • Ver item
    •   Página inicial
    • Teses & Dissertações
    • Teses & Dissertações
    • Ver item
    JavaScript is disabled for your browser. Some features of this site may not work without it.

    Integração de dados de imagens orbitais de alta resolução e ALS para detecção semi-automática de edificações em áreas urbanas

    Thumbnail
    Visualizar/Abrir
    DISSERTACAO_UFPR_FELIPE_M_M_SILVA_impressa.pdf (8.939Mb)
    Data
    2011
    Autor
    Silva, Felipe Martins Marques da
    Metadata
    Mostrar registro completo
    Resumo
    Resumo: Atualmente, os sensores de alta resolução espacial e radiométrica disponíveis, conseguem adquirir imagens nas quais os objetos na superfície terrestre são representados por pixels submétricos. Aliado a isto, tem-se o LiDAR (Light Detection And Ranging), tecnologia de sensoriamento remoto ativo de varredura a laser, que é capaz de coletar coordenadas tridimensionais de milhares de pontos em um curto intervalo de tempo. A tecnologia LiDAR, quando acoplada a aeronaves, é conhecida por ALS (Airborne LiDAR System), e tem sido utilizada para realizar mapeamentos.A integração destas tecnologias ? imagens orbitais de alta resolução e dados ALS ?é desejada pelo fato de fornecerem dados tais que geram informações complementares. Diante disso, esta pesquisa apresenta uma metodologia para identificação de edificações em ambiente urbano, a partir de imagem de alta resolução GeoEye®e dados ALS, baseada em objetos e classificação com base em regras geradas por meio de árvores de decisão. A análise com base em objetos vem da possibilidade de segmentar a imagem, ou seja, dividi-la em regiões espectralmente homogêneas. Com os dados ALS foram gerados o Modelo Digital de Terreno (MDT), o Modelo Digital de Superfície (MDS), e o Modelo Digital de Superfície Normalizado (MDSn), da área de estudo. A partir do MDS e da imagem original obteve-se a imagem ortorretificada que, tendo o MDSn como banda adicional, foi segmentada pelo segmentador FLSA (Full-Lambda Schedule Algorithm). A partir de uma imagem menor, chamada imagem base, foram coletadas amostras de segmentos de classe, que foram utilizadas em uma árvore de decisão para gerar regras de identificação. Estas regras foram aplicadas em (extrapoladas para) uma imagem maior, chamada de imagem teste, de modo a identificar as demais edificações. Nesta pesquisa verifica-se a capacidade de integração dos dados, os atributos que mais contribuem para a identificação de edificações e a metodologia de extrapolação das regras de identificação.
     
    Abstract: Nowadays, the high spatial and radiometric resolutions sensors available are able to get images where objects on the surface are represented by submeter pixels. Along with those sensors, there is the LiDAR (Light Detection And Ranging), an active remote sensing technology able to collect three-dimensional coordinates of thousands of points in a short period of time. This technology, when embedded in aircrafts, is known as ALS (Airborne Laser Scanner), and it has been used to perform mappings. The integration of those technologies ƒ{ high-resolution satellite images and ALS data ƒ{ is desirable because they provide datasets such that are able to generate complementary information. Therefore, this research presents a methodology to identify buildings in an urban area using a GeoEyeR image and ALS data, using an object-based analysis and rule-based classification generated by decision trees. The object-based analysis is possible after the image segmentation, that is, after dividing the image in spectrally homogeneous areas. With the ALS data it was created the Digital Terrain Model (DTM), the Digital Surface Model (DSM), and the Normalized Digital Surface Model (NDSM). By means of the DSM and the original image was made the orthorectified image. Adding the NDSM as a band of the orthorectified image, it was made the image segmentation. In a smaller image, called base image, it was collected samples of the segments in each class, then those samples were used in a decision tree to generate the identification rules. This rules were applied in (extrapolated to) a larger image, called test image, in order to classify another buildings. In this research it was verified the capability of data integration, the attributes that most contribute to identify buildings in urban areas, and the methodology of extrapolating the identification rules.
     
    URI
    https://hdl.handle.net/1884/26634
    Collections
    • Teses & Dissertações [10894]

    DSpace software copyright © 2002-2022  LYRASIS
    Entre em contato | Deixe sua opinião
    Theme by 
    Atmire NV
     

     

    Navegar

    Todo o repositórioComunidades e ColeçõesPor data do documentoAutoresTítulosAssuntosTipoEsta coleçãoPor data do documentoAutoresTítulosAssuntosTipo

    Minha conta

    EntrarCadastro

    Estatística

    Ver as estatísticas de uso

    DSpace software copyright © 2002-2022  LYRASIS
    Entre em contato | Deixe sua opinião
    Theme by 
    Atmire NV