Techniques Used for Geospatial Big Data Collection, Storage and Analysis

Abstract

Nowadays data are generated by different sources, at an incredible rate, and the traditional approaches for their collection, storage and analysis are not suitable. Big Data is analyzed and used by state institutions, business environment, transportation, health, communications, banking system, utilities, defense and other components of modern society in order to support their decisions as well as the human activities. Geospatial data is an important component of Big Data and aerial/satellite images offer a lot of details about the environment, events and their evolution in time. This paper presents the significant techniques used in three main stages of Geospatial Big Data lifecycle, namely data collection, storage and analysis. Geospatial data collections are mainly executed by using web crawlers in order to find meaningful data and, during this stage some preprocessing operations can be done (standardization, completion and integration). Cloud storage and distributed file systems are widely used for Geospatial Big Data storage and new types of non-relational and relational databases are developed. The very challenging aspects for Big Data analysis are related to feature identification and extraction from aerial or satellite images, using feature-based extraction and deep learning algorithms.