Distributed Processing of Projections of Large Datasets: A Preliminary Study
Modern information needs have resulted in very large amounts of data being used in geographic information systems. Problems arise when trying to project these data in a reasonable amount of time and accuracy, however. Current single-threaded methods can suffer from two problems: fast projection with poor accuracy, or accurate projection with long processing time. A possible solution may be to combine accurate interpolation methods and distributed processing algorithms to quickly and accurately convert digital geospatial data between coordinate systems. Modern technology has made it possible to construct systems, such as Beowulf clusters, for a low cost and provide access to supercomputer-class technology. Combining these techniques may result in the ability to use large amounts of geographic data in time-critical situations.
Citation Information
Publication Year | 2004 |
---|---|
Title | Distributed Processing of Projections of Large Datasets: A Preliminary Study |
DOI | 10.3133/ofr2003117 |
Authors | Brian G. Maddox |
Publication Type | Report |
Publication Subtype | USGS Numbered Series |
Series Title | Open-File Report |
Series Number | 2003-117 |
Index ID | ofr2003117 |
Record Source | USGS Publications Warehouse |
USGS Organization | Mid-Continent Mapping Center |