A carregar...
Publicação
Scalable analysis of multitemporal images using an array database
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 1.65 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
Massive archives of earth observation data are now available and the size of this data is increasing at a tremendous rate. This data is a very important resource and has a variety of applications including monitoring change, forestry application, agricultural application and urban planning. At the same time, they also possess challenge of storage, management, and high computational needs. In this study SciDB, an array-based database is used to store, manage and process multitemporal satellite imagery. The major aim of this study is to investigate the performance of SciDB based scalable solution to run arithmetic operation, simple time series analysis and complex time series analysis on multitemporal satellite imagery. This study provides better insight of SciDB architecture and provides suggestions for better performance in SciDB for remote sensing jobs. The research also compared the performance of time series analysis on SciDB array with file-based analysis using multicore parallelization (Using „Parallel‟ Package of R). It is found that SciDB provides a faster solution for time series analysis. However, SciDB might not be the best solution if the data size is smaller. Also, relative immaturity of SciDB and limited inherent support of remote sensing operations increases effort for the scientist to develop SciDB based solution. Nevertheless, SciDB has the potential to meet the ever increasing storage, management and computational need of big remote sensing data.
Descrição
Dissertation submitted in partial fulfilment of the requirements for the degree of Master of Science in Geospatial Technologies
Palavras-chave
Array Database SciDB High Performance Computing Remote Sensing Multitemporal Images