A carregar...
Publicação
Implementação em interface Web do processo de aquisição e estruturação de dados multitemporais obtidos pelo programa de observação da Terra Landsat
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 1.43 MB | Adobe PDF |
Orientador(es)
Resumo(s)
As sucessivas gerações de satélites LANDSAT, geridas por um consórcio que integra a NASA e a USGS, registaram um enorme acervo de imagens da Terra; registos esses que se iniciaram em 1972 e têm decorrido até aos nossos dias. Esse acervo, estruturado em bases de dados, pode ser inquirido e as imagens podem ser descarregadas sem custo. Contudo, todo o processo de aquisição que compreende a inquirição à base de dados e download - é relativamente complexo devido ao facto da plataforma da USGS conter uma enorme densidade de informação. Esta dissertação tem como objectivo principal a concepção e desenvolvimento de uma plataforma WEB, designada GeoRef, para aquisição de imagens de satélite. O GeoRef tem também implementadas algumas funções de processamento e de cálculo sobre imagens LANDSAT, nomeadamente dos sensores MSS,TM,ETM e OLI. Todo o processo de aquisição foi revisto de forma a simplificar a pesquisa, assim como foram operacionalizadas ‘rotinas’ que optimizam a execução de tarefas frequentes. Em média, o tempo necessário para a plataforma produzir resultados, que eram obtidos de forma manual e algumas vêzes em softwares diversos, é de 15 a 20 minutos. Os processos optimizados foram implementados com a integração de software R e linguagem PHP, facto que permitiu a construção de um modelo de utilização sobre os packages do software R, abrindo um variadíssimo leque de opções que poderão a vir integrar a plataforma. Na fase de cálculo foi utilizado o software FRAGSTATS (Spatial Pattern Analysis Program for Categorical Maps) em modo Batch utilizando o mesmo método de integração com R.
Successive generations of LANDSAT satellites, managed by a consortium including NASA and USGS, have recorded a huge collection of Earth surface images – this recording having began in 1972 and continuing until this day. This wealth of data, structured in databases, can be queried and the images can be downloaded without any cost. However, this acquisition process - comprising the querying to the database itself and data downloading - is relatively complex due to the fact that the USGS platform contains a huge density of information. The main aim of this thesis’ was to design and develop a web platform called GEOREF for acquiring these satellite images. GEOREF also implements some processing and calculation functionalities upon Landsat data, namely data coming from the MSS, TM, ETM and OLI sensors. The whole data acquisition process has been tuned in order to simplify the search, and some ‘routines’ were designed which optimize the performance of common tasks. On average, the time required for the platform to produce results is 15 to 20 minutes -- which compares with many hours when this data was obtained manually and frequently using disparate software’s. These optimized processes were implemented by integrating ‘R’ software with PHP language, which also allowed the construction of a model of usage for packages of the ‘R’ software, thus opening a wide variety of options that may later be added into this platform. For the calculations phase, the FRAGSTATS software (“Spatial Pattern Analysis Program for Categorical Maps”) was used in batch mode, resorting to the same method of integration as with ‘R’.
Successive generations of LANDSAT satellites, managed by a consortium including NASA and USGS, have recorded a huge collection of Earth surface images – this recording having began in 1972 and continuing until this day. This wealth of data, structured in databases, can be queried and the images can be downloaded without any cost. However, this acquisition process - comprising the querying to the database itself and data downloading - is relatively complex due to the fact that the USGS platform contains a huge density of information. The main aim of this thesis’ was to design and develop a web platform called GEOREF for acquiring these satellite images. GEOREF also implements some processing and calculation functionalities upon Landsat data, namely data coming from the MSS, TM, ETM and OLI sensors. The whole data acquisition process has been tuned in order to simplify the search, and some ‘routines’ were designed which optimize the performance of common tasks. On average, the time required for the platform to produce results is 15 to 20 minutes -- which compares with many hours when this data was obtained manually and frequently using disparate software’s. These optimized processes were implemented by integrating ‘R’ software with PHP language, which also allowed the construction of a model of usage for packages of the ‘R’ software, thus opening a wide variety of options that may later be added into this platform. For the calculations phase, the FRAGSTATS software (“Spatial Pattern Analysis Program for Categorical Maps”) was used in batch mode, resorting to the same method of integration as with ‘R’.
Descrição
Palavras-chave
Detecção Remota software R GeoRef Imagens Landsat Remote Sensing R software Landsat images