Modern algorithms have estimated there are about 3 trillion trees on Earth, but a more specific analysis requires new methodologies.
Researchers at the U.S. National Aeronautics and Space Administration (NASA) Goddard Space Flight Center and the University of Minnesota have proposed a tree-counting project that combines better algorithms with less costly and faster data processing. The latter would utilize a mix of NASA Goddard's private cloud, as well as the public cloud by way of Amazon Web Services (AWS).
The researchers tested the algorithms in sub-Sarahan Africa, because of the lack of deforestation. The project used images from four commercial satellites, but for the trial the project leaders concentrated on data captured by only one. The data was organized into 225 one-kilometer-square chunks, and NASA scientists first had to create a mosaic of those images to produce a single seam and then make this "stand up" before deploying the counting algorithm, a process known as orthorectified mosaicing. The data presents as a parallel problem, making it suitable for parallel processing on general-purpose server clusters, or processing via cloud services. After mosaicing, the data was pre-staged in AWS and processed in 200 instances using AWS spot pricing.
From The Register (UK)
View Full Article
Abstracts Copyright © 2015 Information Inc., Bethesda, Maryland, USA
No entries found