Improvement of satellite image classification: Approach based on Hadoop/MapReduce

Abstract

The revolution of technologies (social networks, smartphones, GPS and Remote Sensing Image) increase the volume of informations wich makes humanity in new need “Storage of huge volume of data”. the traditional strategy to store data become problem for humanity and then this need build new art to resolve this problems the “Spatial Big Data (SBD)” SBD store proncipally three types of data:vector data, raster data and network data. The complexity and nature of spatial databases make them ideal for applying parallel processing. This also emphasizes the need for developing new efficient geospatial analytic for analyzing spatial big data. So, we review the most used spatial data algorithms attracting human interests especially when the amount of satellite images continues to grow as more information becomes available. In this context, we propose a system based on Hadoop an open source system that implements the MapReduce programming model and that can improve the classification of large scale remote sensing image and benefit the power of spatial big data concept.

6 Figures and Tables

Cite this paper

@article{Chebbi2016ImprovementOS, title={Improvement of satellite image classification: Approach based on Hadoop/MapReduce}, author={Imen Chebbi and Wadii Boulila and Imed Riadh Farah}, journal={2016 2nd International Conference on Advanced Technologies for Signal and Image Processing (ATSIP)}, year={2016}, pages={31-34} }