GPU-specific reformulations of image compression algorithms

D 2012

GPU-specific reformulations of image compression algorithms

MATELA, Jiří, Petr HOLUB, Martin JIRMAN and Martin ŠROM

Basic information

Original name

GPU-specific reformulations of image compression algorithms

Authors

MATELA, Jiří (203 Czech Republic, guarantor, belonging to the institution), Petr HOLUB (203 Czech Republic, belonging to the institution), Martin JIRMAN (203 Czech Republic) and Martin ŠROM (203 Czech Republic)

Edition

8499. vyd. BELLINGHAM, WA 98227-0010 USA, Proceedings of Applications of Digital Image Processing XXXV, p. nestránkováno, 8 pp. 2012

Publisher

SPIE-INT SOC OPTICAL ENGINEERING

Other information

Language

English

Type of outcome

Stať ve sborníku

Field of Study

10201 Computer sciences, information science, bioinformatics

Country of publisher

United States of America

Confidentiality degree

není předmětem státního či obchodního tajemství

Publication form

electronic version available online

RIV identification code

RIV/00216224:14330/12:00057518

Organization unit

Faculty of Informatics

ISBN

978-0-8194-9216-6

ISSN

DOI

http://dx.doi.org/10.1117/12.929971

UT WoS

000312200600040

Keywords in English

GPU; parallel; reformulation; JPEG; JPEG2000; Context Modeling; Arithmetic coding; MQ-Coder; Huffman coding

Abstract

V originále

Image compressions have a number of applications in various fields where the processing throughput and/or latency is a crucial attribute and the main limitation with state of the art implementations of compression algorithms. At the same time the contemporary GPUs provide a tremendous processing power applicable to the image compression acceleration but it calls for a specific algorithm design. We discuss the key components of successful GPU algorithm design and demonstrate this on JPEG2000 compression chain, which contains several types of algorithms: from DWT which is inherently well suited to GPU, through context modeling requiring reformulation in order to perform well on GPU, to arithmetic coding which does not fit the paradigm well but can be optimized to perform faster than CPU versions. Performance evaluation of the optimized JPEG2000 chain will be used to demonstrate the importance of various aspects of GPU programming, especially with respect to real-time applications.

Links

GAP202/12/0306, research and development project

Name: Dyschnet - Dynamické plánování a rozvrhování výpočetních a síťových zdrojů (Acronym: Dyschnet)

Investor: Czech Science Foundation

GD102/09/H042, research and development project

Name: Matematické a inženýrské metody pro vývoj spolehlivých a bezpečných paralelních a distribuovaných počítačových systémů

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

Citovat

MATELA, Jiří, Petr HOLUB, Martin JIRMAN and Martin ŠROM. GPU-specific reformulations of image compression algorithms. Online. In Tescher, AG. Proceedings of Applications of Digital Image Processing XXXV. 8499th ed. BELLINGHAM, WA 98227-0010 USA: SPIE-INT SOC OPTICAL ENGINEERING, 2012, p. nestránkováno, 8 pp. ISBN 978-0-8194-9216-6. Available from: https://dx.doi.org/10.1117/12.929971.

@inproceedings{989228,
   author = {Matela, Jiří and Holub, Petr and Jirman, Martin and Šrom, Martin},
   address = {BELLINGHAM, WA 98227-0010 USA},
   booktitle = {Proceedings of Applications of Digital Image Processing XXXV},
   doi = {http://dx.doi.org/10.1117/12.929971},
   edition = {8499},
   editor = {Tescher, AG},
   keywords = {GPU; parallel; reformulation; JPEG; JPEG2000; Context Modeling; Arithmetic coding; MQ-Coder; Huffman coding},
   howpublished = {elektronická verze "online"},
   language = {eng},
   location = {BELLINGHAM, WA 98227-0010 USA},
   isbn = {978-0-8194-9216-6},
   pages = {nestránkováno},
   publisher = {SPIE-INT SOC OPTICAL ENGINEERING},
   title = {GPU-specific reformulations of image compression algorithms},
   year = {2012}
}

TY  - JOUR
ID  - 989228
AU  - Matela, Jiří - Holub, Petr - Jirman, Martin - Šrom, Martin
PY  - 2012
TI  - GPU-specific reformulations of image compression algorithms
PB  - SPIE-INT SOC OPTICAL ENGINEERING
CY  - BELLINGHAM, WA 98227-0010 USA
SN  - 9780819492166
KW  - GPU
KW  - parallel
KW  - reformulation
KW  - JPEG
KW  - JPEG2000
KW  - Context Modeling
KW  - Arithmetic coding
KW  - MQ-Coder
KW  - Huffman coding
N2  - Image compressions have a number of applications in various fields where the processing throughput and/or latency is a crucial attribute and the main limitation with state of the art implementations of compression algorithms. At the same time the contemporary GPUs provide a tremendous processing power applicable to the image compression acceleration but it calls for a specific algorithm design. We discuss the key components of successful GPU algorithm design and demonstrate this on JPEG2000 compression chain, which contains several types of algorithms: from DWT which is inherently well suited to GPU, through context modeling requiring reformulation in order to perform well on GPU, to arithmetic coding which does not fit the paradigm well but can be optimized to perform faster than CPU versions. Performance evaluation of the optimized JPEG2000 chain will be used to demonstrate the importance of various aspects of GPU programming, especially with respect to real-time applications.
ER  -

MATELA, Jiří, Petr HOLUB, Martin JIRMAN and Martin ŠROM. GPU-specific reformulations of image compression algorithms. Online. In Tescher, AG. \textit{Proceedings of Applications of Digital Image Processing XXXV}. 8499th ed. BELLINGHAM, WA 98227-0010 USA: SPIE-INT SOC OPTICAL ENGINEERING, 2012, p.~nestránkováno, 8 pp. ISBN~978-0-8194-9216-6. Available from: https://dx.doi.org/10.1117/12.929971.

Detailed Information on Publication Record