Please use this identifier to cite or link to this item:
|Title:||Optimization of the Processing of Time series of different Satellite Imagery|
|Authors:||DE JAGER Alfred|
|Citation:||Dataflow from Space to Earth p. 1-11|
|JRC Publication N°:||JRC64743|
|Type:||Contributions to Conferences|
|Abstract:||With the increase in availability of space born data, analysing large amounts of high resolution data of significant time series evolved to a common practice. From a computational perspective the analysis consists, in many cases, of comparing the value of stacks of pixels at the same location at different moments in time. Thus the development of a certain band (reflection) can be measured and anomalies can detected. The rather simple computations involved in stacking pixels in spaceborn imagery are hindered by differences in georeferencing between the various images. Before pixels in stack of images can be compared the imagery has to be referenced to a common grid allowing to make valid cross analysis. We defined a world reference grid based on decimal degrees and natural divisions of decimal degrees. Using this Grid we were able to separate the computational intensive georeferencing from the actual analysing of huge amounts of imagery data. The reference grid is stored in a spatial relational database and location dependend identifiers are assigned to every grid cell. These identifiers, so-called C-SQUARES, allow for quick switching between various cell sizes (resolutions). Subsequently images are georeferenced and every pixel is resampled against the static reference. Once the most appropriate value for a specific image overlaid by one or more grid cells is determined the result value is stored in a timeseries referenced by the reference grid cell. Once this precise georeferencing is done, the timeseries can be analysed very quickly. Remaining computations involve mainly the comparison of numbers referring to the same grid cell. In this paper we will give insight in the creation of the reference grid and the savings in storage space and processing time achieved by applying this methodology using MERIS data over Europe and Africa. Specific Web mapping applications showing time series of 10 day composites for 10 years will be shown.|
|JRC Institute:||Institute for Environment and Sustainability|
Files in This Item:
There are no files associated with this item.
Items in repository are protected by copyright, with all rights reserved, unless otherwise indicated.