In this paper we study the distributed system design to handle big geospatial data. We
found this area of research
is very active in terms of review, development and usage of the existing
solutions with a continuous
implementation for specific use cases and application areas such as disaster management environmental
monitoring, earth observation data analysis and distribution, etc. We have
attempted to compare Big Data,
Geospatial Big Data and Geospatial Data to clarify the possible features of differences, compare them in
the term
of storage and processing background for different data representation and tried to collect and categorized the
existing common system solutions. The second part of the paper provided overview about our previous work and
implementation the new framework called IQLib. We have introduced the 3 main modules of the system and
described in brief their technical realization. Data Catalogue module is on higher level of preparation, hence we
could provide details on the data model and data decomposition.
IQLib documentation on data model and data catalogue
is available on Github at
https://github.com/posseidon/iqlib.
We have decided not to publish Data Catalogue module’s source code until it has been reviewed and finalized by
IQmulus project partners. However, the RESTFUL API is available on Heroku cloud infrastructure for all
project partners to test and give feedbacks and suggestions at http://iqlib.herokuapp.com. In our future work we
are going to focus on further development of the framework along with the processing executables and
experimental benchmarking of processing time. Next development phase is the
implementation of Data
Catalogue module and installation for testing phase. Future directions are the Processing module to be
implemented together with GIS raster processing scripts for further analysis of open geodata (such as Sentinel 2
satellite imagery).
Dostları ilə paylaş: