Parallel - Scalable parallel image geoprocessing techniques for high-throughput/high-speed product generation
The increasingly operational use of geospatial imagery, through initiatives such as integrated systems for environmental monitoring and security, is causing a huge increase in the volume of data to be processed. At the same time, turnaround times for GIS-ready products must become significantly shorter to meet user requirements for service timeliness. Rectification of EO data – the process of resampling image pixels from the sensor’s coordinate system into a known geographic coordinate system – remains one of the most computationally demanding steps in this process and is a potential bottleneck to overall service delivery that must be addressed in the next generation of image production systems.
Parallel processing techniques applied to image rectification have the potential to meet critical performance needs. Technologies for parallel processing are already well established in large business systems and such techniques have previously been applied to geoprocessing of satellite image quicklooks within the framework of the WMS Image Server developed by Spacemetric for the European Space Agency. The subject of the PARALLEL project is the application of such techniques in a robust and scalable high-performance solution for georectification of full-resolution imagery.
The project will capitalise upon the specific characteristics of Application Servers and multi-CPU/GRID computing to provide a generic approach to parallel georectification of imagery.