GPU Optimization of Convolution for Large 3-D Real Images

D 2012

GPU Optimization of Convolution for Large 3-D Real Images

KARAS, Pavel, David SVOBODA and Pavel ZEMČÍK

Basic information

Original name

GPU Optimization of Convolution for Large 3-D Real Images

Authors

KARAS, Pavel (203 Czech Republic, guarantor, belonging to the institution), David SVOBODA (203 Czech Republic, belonging to the institution) and Pavel ZEMČÍK (203 Czech Republic)

Edition

Neuveden, Proceedings of the International Conference on Advanced Concepts for Intelligent Vision Systems (ACIVS’12), p. 59-71, 13 pp. 2012

Publisher

Springer Berlin / Heidelberg

Other information

Language

English

Type of outcome

Stať ve sborníku

Field of Study

10201 Computer sciences, information science, bioinformatics

Country of publisher

Germany

Confidentiality degree

není předmětem státního či obchodního tajemství

Publication form

printed version "print"

References:

URL

Impact factor

Impact factor: 0.402 in 2005

RIV identification code

RIV/00216224:14330/12:00057444

Organization unit

Faculty of Informatics

ISBN

978-3-642-33139-8

ISSN

DOI

http://dx.doi.org/10.1007/978-3-642-33140-4_6

Keywords in English

gpu; convolution; 3-D; image processing

Abstract

V originále

In this paper, we propose a method for computing convolution of large 3-D images with respect to real signals. The convolution is performed in a frequency domain using a convolution theorem. Due to properties of real signals, the algorithm can be optimized so that both time and the memory consumption are halved when compared to complex signals of the same size. Convolution is decomposed in a frequency domain using the decimation in frequency (DIF) algorithm. The algorithm is accelerated on a graphics hardware by means of the CUDA parallel computing model, achieving up to 10x speedup with a single GPU over an optimized implementation on a quad-core CPU.

Links

GBP302/12/G157, research and development project

Name: Dynamika a organizace chromosomů během buněčného cyklu a při diferenciaci v normě a patologii

Investor: Czech Science Foundation

MUNI/A/0914/2009, interní kód MU

Name: Rozsáhlé výpočetní systémy: modely, aplikace a verifikace (Acronym: SV-FI MAV)

Investor: Masaryk University, Category A

2B06052, research and development project

Name: Vytipování markerů, screening a časná diagnostika nádorových onemocnění pomocí vysoce automatizovaného zpracování multidimenzionálních biomedicínských obrazů (Acronym: Biomarker)

Investor: Ministry of Education, Youth and Sports of the CR, Determination of markers, screening and early diagnostics of cancer diseases using highly automated processing of multidimensional biomedical images

Citovat

KARAS, Pavel, David SVOBODA and Pavel ZEMČÍK. GPU Optimization of Convolution for Large 3-D Real Images. In Blanc-Talon, Jacques and Philips, Wilfried and Popescu, Dan and Scheunders, Paul and Zemcík, Pavel. Proceedings of the International Conference on Advanced Concepts for Intelligent Vision Systems (ACIVS’12). Neuveden: Springer Berlin / Heidelberg, 2012, p. 59-71. ISBN 978-3-642-33139-8. Available from: https://dx.doi.org/10.1007/978-3-642-33140-4_6.

@inproceedings{984851,
   author = {Karas, Pavel and Svoboda, David and Zemčík, Pavel},
   address = {Neuveden},
   booktitle = {Proceedings of the International Conference on Advanced Concepts for Intelligent Vision Systems (ACIVS’12)},
   doi = {http://dx.doi.org/10.1007/978-3-642-33140-4_6},
   editor = {Blanc-Talon, Jacques and Philips, Wilfried and Popescu, Dan and Scheunders, Paul and Zemcík, Pavel},
   keywords = {gpu; convolution; 3-D; image processing},
   howpublished = {tištěná verze "print"},
   language = {eng},
   location = {Neuveden},
   isbn = {978-3-642-33139-8},
   pages = {59-71},
   publisher = {Springer Berlin / Heidelberg},
   title = {GPU Optimization of Convolution for Large 3-D Real Images},
   url = {http://dx.doi.org/10.1007/978-3-642-33140-4_6},
   year = {2012}
}

TY  - JOUR
ID  - 984851
AU  - Karas, Pavel - Svoboda, David - Zemčík, Pavel
PY  - 2012
TI  - GPU Optimization of Convolution for Large 3-D Real Images
PB  - Springer Berlin / Heidelberg
CY  - Neuveden
SN  - 9783642331398
KW  - gpu
KW  - convolution
KW  - 3-D
KW  - image processing
UR  - http://dx.doi.org/10.1007/978-3-642-33140-4_6
N2  - In this paper, we propose a method for computing convolution of large 3-D images with respect to real signals. The convolution is performed in a frequency domain using a convolution theorem. Due to properties of real signals, the algorithm can be optimized so that both time and the memory consumption are halved when compared to complex signals of the same size. Convolution is decomposed in a frequency domain using the decimation in frequency (DIF) algorithm. The algorithm is accelerated on a graphics hardware by means of the CUDA parallel computing model, achieving up to 10x speedup with a single GPU over an optimized implementation on a quad-core CPU.
ER  -

KARAS, Pavel, David SVOBODA and Pavel ZEMČÍK. GPU Optimization of Convolution for Large 3-D Real Images. In Blanc-Talon, Jacques and Philips, Wilfried and Popescu, Dan and Scheunders, Paul and Zemcík, Pavel. \textit{Proceedings of the International Conference on Advanced Concepts for Intelligent Vision Systems (ACIVS’12)}. Neuveden: Springer Berlin / Heidelberg, 2012, p.~59-71. ISBN~978-3-642-33139-8. Available from: https://dx.doi.org/10.1007/978-3-642-33140-4\_{}6.

Detailed Information on Publication Record