Research Article
GRIMS: A Scalable Management and Storage System for Massive Remote Sensing Images
@INPROCEEDINGS{10.4108/ICST.INFOSCALE2008.3516, author={Liang ZHAO and Luo CHEN and Ning JING and Huaiyu ZUO}, title={GRIMS: A Scalable Management and Storage System for Massive Remote Sensing Images}, proceedings={3rd International ICST Conference on Scalable Information Systems}, publisher={ICST}, proceedings_a={INFOSCALE}, year={2010}, month={5}, keywords={RS images GRIMS Management Partition Storage D-MaRISS}, doi={10.4108/ICST.INFOSCALE2008.3516} }- Liang ZHAO
Luo CHEN
Ning JING
Huaiyu ZUO
Year: 2010
GRIMS: A Scalable Management and Storage System for Massive Remote Sensing Images
INFOSCALE
ICST
DOI: 10.4108/ICST.INFOSCALE2008.3516
Abstract
Remote sensing images are important in ecological, geographical and military applications. With the rapid growing volume of remote sensing images, how to manage and store the massive remote sensing images is becoming a must-be-solved problem. We build a scalable storage and management system for massive remote sensing images aiming to store global remote sensing images -- Global Remote Sensing Images Management and Storage system (GRIMS). In GRIMS, we propose a tile pyramid model -- Plate Carree Projection Grid Quad-Tree (PCPGQT) to spatially partition the high resolution images and build a double tower (DT) index for the big image data file. Meanwhile, we analyze the fragment problem in image mosaic. Through fragment collection, huge disk spaces could be saved. By using the open source software HADOOP, we realize a distributed storage system to store massive image data file. Experiments show that our tile pyramid model is suitable to support our system purpose, and the distributed storage system is highly efficient.
