| Article ID | Journal | Published Year | Pages | File Type |
|---|---|---|---|---|
| 1784992 | Infrared Physics & Technology | 2007 | 4 Pages |
Abstract
Massively parallel computational paradigms can mitigate many issues associated with the analysis of large and complex remotely sensed data sets. Recently, the Beowulf cluster has emerged as the most attractive, massively parallel architecture due to its low cost and high performance. Whereas most Beowulf designs have emphasized numerical modeling applications, the Parallel Image Processing Environment (PIPE) specifically addresses the unique requirements of remote sensing applications. Automated, parallelization of user-defined analyses is fully supported. A neural network application, applied to Along Track Scanning Radiometer-2 (ATSR-2) data shows the advantages and performance characteristics of PIPE.
Related Topics
Physical Sciences and Engineering
Physics and Astronomy
Atomic and Molecular Physics, and Optics
Authors
J. Simpson, T. McIntire, J. Berg, Y. Tsou,