Big Earth Observation Data Integration in Remote Sensing Based on a Distributed Spatial Framework

Abstract

The arrival of the era of big data for Earth observation (EO) indicates that traditional data management models have been unable to meet the needs of remote sensing data in big data environments. With the launch of the first remote sensing satellite, the volume of remote sensing data has also been increasing, and traditional data storage methods have been unable to ensure the efficient management of large amounts of remote sensing data. Therefore, a professional remote sensing big data integration method is sorely needed. In recent years, the emergence of some new technical methods has provided effective solutions for multi-source remote sensing data integration. This paper proposes a multi-source remote sensing data integration framework based on a distributed management model. In this framework, the multi-source remote sensing data are partitioned by the proposed spatial segmentation indexing (SSI) model through spatial grid segmentation. The designed complete information description system, based on International Organization for Standardization (ISO) 19115, can explain multi-source remote sensing data in detail. Then, the distributed storage method of data based on MongoDB is used to store multi-source remote sensing data. The distributed storage method is physically based on the sharding mechanism of the MongoDB database, and it can provide advantages for the security and performance of the preservation of remote sensing data. Finally, several experiments have been designed to test the performance of this framework in integrating multi-source remote sensing data. The results show that the storage and retrieval performance of the distributed remote sensing data integration framework proposed in this paper is superior. At the same time, the grid level of the SSI model proposed in this paper also has an important impact on the storage efficiency of remote sensing data. Therefore, the remote storage data integration framework, based on distributed storage, can provide new technical support and development prospects for big EO data.

Big Data Integration Solutions in Organizations: A Domain-Specific Analysis

“…Cheng et al proposed a remote sensing data management system [37]. This system is distributed multisource and followed the MongoDB model.…”

An improved tile-based scalable distributed management model of massive high-resolution satellite images

“…The most common method to organize image data blocks is to build up a hierarchic and multiscale pyramid model with a global sub-division grid image. Therefore, end-users could easily reach the image blocks of particular spatial regions and specific resolution levels as needed [30]. Generally, most RS images are composed of multiple bands.…”

Multitemporal Mosaicing for Sentinel-3/FLEX Derived Level-2 Product Composites

“…The first step of the proposed framework aims at collecting the corresponding OLCI L2 products of interest for generating the output multi-temporal mosaic. It is important to highlight that the task of generating global composites usually demands dealing with vast amounts of operational RS data [41], which logically raises some technical challenges (in terms of data acquisition, storage and automatization) that this initial step tries to cope with. More specifically, we make use of the available Copernicus data provision services in the following way.…”