Highly accurate distortion estimation for JPEG2000 through pdf-based estimators

Francesc Aulí-Llinàs, Michael W Marcellin, Joan Serra-Sagristà

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

Abstract

Distortion estimation techniques are often employed in bitplane coding engines to minimize the computational load, or the memory requirements, of the encoder. A common approach is to determine distortion estimators that approximate the mean squared error decreases when data are successively coded and transmitted. Such estimators usually assume that coefficients are uniformly distributed in the quantization interval. Even though this assumption simplifies estimation, it does not exactly correspond with the nature of the signal. This work introduces new distortion estimators determined through a precise approximation of the coefficient's distribution within the quantization intervals. Experimental results obtained when our estimators are used for the post-compression rate-distortion optimization process of JPEG2000 suggest that they are able to approximate distortion with very high accuracy.

Original languageEnglish (US)
Title of host publicationData Compression Conference Proceedings
Pages391-400
Number of pages10
StatePublished - 2009
Event2009 Data Compression Conference, DCC 2009 - Snowbird, UT, United States
Duration: Mar 16 2010Mar 18 2010

Other

Other2009 Data Compression Conference, DCC 2009
CountryUnited States
CitySnowbird, UT
Period3/16/103/18/10

Fingerprint

Engines
Data storage equipment

ASJC Scopus subject areas

  • Computer Networks and Communications

Cite this

Aulí-Llinàs, F., Marcellin, M. W., & Serra-Sagristà, J. (2009). Highly accurate distortion estimation for JPEG2000 through pdf-based estimators. In Data Compression Conference Proceedings (pp. 391-400). [4976483]

Highly accurate distortion estimation for JPEG2000 through pdf-based estimators. / Aulí-Llinàs, Francesc; Marcellin, Michael W; Serra-Sagristà, Joan.

Data Compression Conference Proceedings. 2009. p. 391-400 4976483.

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

Aulí-Llinàs, F, Marcellin, MW & Serra-Sagristà, J 2009, Highly accurate distortion estimation for JPEG2000 through pdf-based estimators. in Data Compression Conference Proceedings., 4976483, pp. 391-400, 2009 Data Compression Conference, DCC 2009, Snowbird, UT, United States, 3/16/10.
Aulí-Llinàs F, Marcellin MW, Serra-Sagristà J. Highly accurate distortion estimation for JPEG2000 through pdf-based estimators. In Data Compression Conference Proceedings. 2009. p. 391-400. 4976483
Aulí-Llinàs, Francesc ; Marcellin, Michael W ; Serra-Sagristà, Joan. / Highly accurate distortion estimation for JPEG2000 through pdf-based estimators. Data Compression Conference Proceedings. 2009. pp. 391-400
@inproceedings{dfee95e612d240078ff686cea7f1b268,
title = "Highly accurate distortion estimation for JPEG2000 through pdf-based estimators",
abstract = "Distortion estimation techniques are often employed in bitplane coding engines to minimize the computational load, or the memory requirements, of the encoder. A common approach is to determine distortion estimators that approximate the mean squared error decreases when data are successively coded and transmitted. Such estimators usually assume that coefficients are uniformly distributed in the quantization interval. Even though this assumption simplifies estimation, it does not exactly correspond with the nature of the signal. This work introduces new distortion estimators determined through a precise approximation of the coefficient's distribution within the quantization intervals. Experimental results obtained when our estimators are used for the post-compression rate-distortion optimization process of JPEG2000 suggest that they are able to approximate distortion with very high accuracy.",
author = "Francesc Aul{\'i}-Llin{\`a}s and Marcellin, {Michael W} and Joan Serra-Sagrist{\`a}",
year = "2009",
language = "English (US)",
isbn = "9780769535920",
pages = "391--400",
booktitle = "Data Compression Conference Proceedings",

}

TY - GEN

T1 - Highly accurate distortion estimation for JPEG2000 through pdf-based estimators

AU - Aulí-Llinàs, Francesc

AU - Marcellin, Michael W

AU - Serra-Sagristà, Joan

PY - 2009

Y1 - 2009

N2 - Distortion estimation techniques are often employed in bitplane coding engines to minimize the computational load, or the memory requirements, of the encoder. A common approach is to determine distortion estimators that approximate the mean squared error decreases when data are successively coded and transmitted. Such estimators usually assume that coefficients are uniformly distributed in the quantization interval. Even though this assumption simplifies estimation, it does not exactly correspond with the nature of the signal. This work introduces new distortion estimators determined through a precise approximation of the coefficient's distribution within the quantization intervals. Experimental results obtained when our estimators are used for the post-compression rate-distortion optimization process of JPEG2000 suggest that they are able to approximate distortion with very high accuracy.

AB - Distortion estimation techniques are often employed in bitplane coding engines to minimize the computational load, or the memory requirements, of the encoder. A common approach is to determine distortion estimators that approximate the mean squared error decreases when data are successively coded and transmitted. Such estimators usually assume that coefficients are uniformly distributed in the quantization interval. Even though this assumption simplifies estimation, it does not exactly correspond with the nature of the signal. This work introduces new distortion estimators determined through a precise approximation of the coefficient's distribution within the quantization intervals. Experimental results obtained when our estimators are used for the post-compression rate-distortion optimization process of JPEG2000 suggest that they are able to approximate distortion with very high accuracy.

UR - http://www.scopus.com/inward/record.url?scp=77957745867&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77957745867&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9780769535920

SP - 391

EP - 400

BT - Data Compression Conference Proceedings

ER -