Title :
MEMCA for satellite and space data: MEMCA [Memory centric analytics] for satellite and space data
Author :
Agca, Muhammed Akif ; Baceski, Emre ; Gokcebag, Serhan
Author_Institution :
HAVELSAN Ankara, Ankara, Turkey
Abstract :
Automated image exploitation algorithms requires both computation power and fast disk access. Ordinary hardware or a sequential processing architecture is far from performing with acceptable performance, while dealing with huge satellite imagery data. In this work, a memory centric analytics solution is proposed for performing querying, graph based operations, machine learning and stream processing effectively. A proof of concept experiment was performed in order to demonstrate the effect of memory centric operations on big data. The initial results show that, memory centric analytics has great potential for imagery exploitation.
Keywords :
data handling; geophysical image processing; learning (artificial intelligence); MEMCA; fast disk access; image exploitation algorithms; imagery exploitation; machine learning; memory centric analytics; memory centric operations; satellite imagery data; sequential processing architecture; space data; stream processing; File systems; Memory architecture; Memory management; Random access memory; Satellites; Streaming media; big data; distributed system; memory centric file system; satellite images; stream processing;
Conference_Titel :
Recent Advances in Space Technologies (RAST), 2015 7th International Conference on
Conference_Location :
Istanbul
Print_ISBN :
978-1-4673-7760-7
DOI :
10.1109/RAST.2015.7208352