Massively-parallel break detection for satellite data

Malte von Mehren, Fabian Gieseke, Jan Verbesselt, Sabina Rosca, Stéphanie Horion, Achim Zeileis

Research output: Chapter in Book/Report/Conference proceedingConference paper

Abstract

The field of remote sensing is nowadays faced with huge amounts of data. While this offers a variety of exciting research opportunities, it also yields significant challenges regarding both computation time and space requirements. In practice, the sheer data volumes render existing approaches too slow for processing and analyzing all the available data. This work aims at accelerating BFAST, one of the state-of-the-art methods for break detection given satellite image time series. In particular, we propose a massively-parallel implementation for BFAST that can effectively make use of modern parallel compute devices such as GPUs. Our experimental evaluation shows that the proposed GPU implementation is up to four orders of magnitude faster than the existing publicly available implementation and up to ten times faster than a corresponding multi-threaded CPU execution. The dramatic decrease in running time renders the analysis of significantly larger datasets possible in seconds or minutes instead of hours or days. We demonstrate the practical benefits of our implementations given both artificial and real datasets.

Original languageEnglish
Title of host publicationProceedings of the 30th International Conference on Scientific and Statistical Database Management
Place of PublicationNew York
PublisherAssociation for Computing Machinery
VolumePart F137913
ISBN (Electronic)9781450365055
DOIs
Publication statusPublished - 9 Jul 2018
Event30th International Conference on Scientific and Statistical Database Management, SSDBM 2018 - Bolzano-Bozen, Italy
Duration: 9 Jul 201811 Jul 2018

Conference

Conference30th International Conference on Scientific and Statistical Database Management, SSDBM 2018
CountryItaly
CityBolzano-Bozen
Period9/07/1811/07/18

Fingerprint

Satellites
Program processors
Time series
Remote sensing
Processing
Graphics processing unit

Cite this

von Mehren, M., Gieseke, F., Verbesselt, J., Rosca, S., Horion, S., & Zeileis, A. (2018). Massively-parallel break detection for satellite data. In Proceedings of the 30th International Conference on Scientific and Statistical Database Management (Vol. Part F137913). [5] New York: Association for Computing Machinery. https://doi.org/10.1145/3221269.3223032
von Mehren, Malte ; Gieseke, Fabian ; Verbesselt, Jan ; Rosca, Sabina ; Horion, Stéphanie ; Zeileis, Achim. / Massively-parallel break detection for satellite data. Proceedings of the 30th International Conference on Scientific and Statistical Database Management. Vol. Part F137913 New York : Association for Computing Machinery, 2018.
@inproceedings{35e3c5ad5c9549819a46e26523c7faf7,
title = "Massively-parallel break detection for satellite data",
abstract = "The field of remote sensing is nowadays faced with huge amounts of data. While this offers a variety of exciting research opportunities, it also yields significant challenges regarding both computation time and space requirements. In practice, the sheer data volumes render existing approaches too slow for processing and analyzing all the available data. This work aims at accelerating BFAST, one of the state-of-the-art methods for break detection given satellite image time series. In particular, we propose a massively-parallel implementation for BFAST that can effectively make use of modern parallel compute devices such as GPUs. Our experimental evaluation shows that the proposed GPU implementation is up to four orders of magnitude faster than the existing publicly available implementation and up to ten times faster than a corresponding multi-threaded CPU execution. The dramatic decrease in running time renders the analysis of significantly larger datasets possible in seconds or minutes instead of hours or days. We demonstrate the practical benefits of our implementations given both artificial and real datasets.",
author = "{von Mehren}, Malte and Fabian Gieseke and Jan Verbesselt and Sabina Rosca and St{\'e}phanie Horion and Achim Zeileis",
year = "2018",
month = "7",
day = "9",
doi = "10.1145/3221269.3223032",
language = "English",
volume = "Part F137913",
booktitle = "Proceedings of the 30th International Conference on Scientific and Statistical Database Management",
publisher = "Association for Computing Machinery",

}

von Mehren, M, Gieseke, F, Verbesselt, J, Rosca, S, Horion, S & Zeileis, A 2018, Massively-parallel break detection for satellite data. in Proceedings of the 30th International Conference on Scientific and Statistical Database Management. vol. Part F137913, 5, Association for Computing Machinery, New York, 30th International Conference on Scientific and Statistical Database Management, SSDBM 2018, Bolzano-Bozen, Italy, 9/07/18. https://doi.org/10.1145/3221269.3223032

Massively-parallel break detection for satellite data. / von Mehren, Malte; Gieseke, Fabian; Verbesselt, Jan; Rosca, Sabina; Horion, Stéphanie; Zeileis, Achim.

Proceedings of the 30th International Conference on Scientific and Statistical Database Management. Vol. Part F137913 New York : Association for Computing Machinery, 2018. 5.

Research output: Chapter in Book/Report/Conference proceedingConference paper

TY - GEN

T1 - Massively-parallel break detection for satellite data

AU - von Mehren, Malte

AU - Gieseke, Fabian

AU - Verbesselt, Jan

AU - Rosca, Sabina

AU - Horion, Stéphanie

AU - Zeileis, Achim

PY - 2018/7/9

Y1 - 2018/7/9

N2 - The field of remote sensing is nowadays faced with huge amounts of data. While this offers a variety of exciting research opportunities, it also yields significant challenges regarding both computation time and space requirements. In practice, the sheer data volumes render existing approaches too slow for processing and analyzing all the available data. This work aims at accelerating BFAST, one of the state-of-the-art methods for break detection given satellite image time series. In particular, we propose a massively-parallel implementation for BFAST that can effectively make use of modern parallel compute devices such as GPUs. Our experimental evaluation shows that the proposed GPU implementation is up to four orders of magnitude faster than the existing publicly available implementation and up to ten times faster than a corresponding multi-threaded CPU execution. The dramatic decrease in running time renders the analysis of significantly larger datasets possible in seconds or minutes instead of hours or days. We demonstrate the practical benefits of our implementations given both artificial and real datasets.

AB - The field of remote sensing is nowadays faced with huge amounts of data. While this offers a variety of exciting research opportunities, it also yields significant challenges regarding both computation time and space requirements. In practice, the sheer data volumes render existing approaches too slow for processing and analyzing all the available data. This work aims at accelerating BFAST, one of the state-of-the-art methods for break detection given satellite image time series. In particular, we propose a massively-parallel implementation for BFAST that can effectively make use of modern parallel compute devices such as GPUs. Our experimental evaluation shows that the proposed GPU implementation is up to four orders of magnitude faster than the existing publicly available implementation and up to ten times faster than a corresponding multi-threaded CPU execution. The dramatic decrease in running time renders the analysis of significantly larger datasets possible in seconds or minutes instead of hours or days. We demonstrate the practical benefits of our implementations given both artificial and real datasets.

U2 - 10.1145/3221269.3223032

DO - 10.1145/3221269.3223032

M3 - Conference paper

VL - Part F137913

BT - Proceedings of the 30th International Conference on Scientific and Statistical Database Management

PB - Association for Computing Machinery

CY - New York

ER -

von Mehren M, Gieseke F, Verbesselt J, Rosca S, Horion S, Zeileis A. Massively-parallel break detection for satellite data. In Proceedings of the 30th International Conference on Scientific and Statistical Database Management. Vol. Part F137913. New York: Association for Computing Machinery. 2018. 5 https://doi.org/10.1145/3221269.3223032