PERFORMANCE OF BLOCK COSINE IMAGE CODING WITH ADAPTIVE QUANTIZATION.

James W. Modestino, Nariman Farvardin, Michael A. Ogrinc

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

Quantizers for block transform image coding systems are typically designed under the assumption of Gaussian statistics for the transform coefficients. While convincing arguments can be provided in support of this approach, empirical evidence is presented demonstrating that, except possibly for the dc term, wide departures from Gaussian behavior can be expected for real-world imagery at typical block sizes. In this paper we describe the performance of a block cosine image coding system with an adaptive quantizer matched to the statistics of the transform coefficients. The adaptive quantizer is based upon a recently developed algorithm which employs a training sequence in the design procedure. At encoding rates of approximately 1 bit/pixel and above, this approach results in significant improvement in reconstructed image quality compared to fixed quantization schemes designed under the Gaussian assumption. For rates much below 1 bit/pixel the relative improvement is negligible.

Original languageEnglish
Pages (from-to)210-217
Number of pages8
JournalIEEE Transactions on Communications
VolumeCOM-33
Issue number3
StatePublished - Mar 1 1985

Fingerprint

Image coding
Pixels
Statistics
Image quality

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Modestino, J. W., Farvardin, N., & Ogrinc, M. A. (1985). PERFORMANCE OF BLOCK COSINE IMAGE CODING WITH ADAPTIVE QUANTIZATION. IEEE Transactions on Communications, COM-33(3), 210-217.

PERFORMANCE OF BLOCK COSINE IMAGE CODING WITH ADAPTIVE QUANTIZATION. / Modestino, James W.; Farvardin, Nariman; Ogrinc, Michael A.

In: IEEE Transactions on Communications, Vol. COM-33, No. 3, 01.03.1985, p. 210-217.

Research output: Contribution to journalArticle

Modestino, JW, Farvardin, N & Ogrinc, MA 1985, 'PERFORMANCE OF BLOCK COSINE IMAGE CODING WITH ADAPTIVE QUANTIZATION.', IEEE Transactions on Communications, vol. COM-33, no. 3, pp. 210-217.
Modestino JW, Farvardin N, Ogrinc MA. PERFORMANCE OF BLOCK COSINE IMAGE CODING WITH ADAPTIVE QUANTIZATION. IEEE Transactions on Communications. 1985 Mar 1;COM-33(3):210-217.
Modestino, James W. ; Farvardin, Nariman ; Ogrinc, Michael A. / PERFORMANCE OF BLOCK COSINE IMAGE CODING WITH ADAPTIVE QUANTIZATION. In: IEEE Transactions on Communications. 1985 ; Vol. COM-33, No. 3. pp. 210-217.
@article{44bf98dc57ab4f32b49e9b3d33349dba,
title = "PERFORMANCE OF BLOCK COSINE IMAGE CODING WITH ADAPTIVE QUANTIZATION.",
abstract = "Quantizers for block transform image coding systems are typically designed under the assumption of Gaussian statistics for the transform coefficients. While convincing arguments can be provided in support of this approach, empirical evidence is presented demonstrating that, except possibly for the dc term, wide departures from Gaussian behavior can be expected for real-world imagery at typical block sizes. In this paper we describe the performance of a block cosine image coding system with an adaptive quantizer matched to the statistics of the transform coefficients. The adaptive quantizer is based upon a recently developed algorithm which employs a training sequence in the design procedure. At encoding rates of approximately 1 bit/pixel and above, this approach results in significant improvement in reconstructed image quality compared to fixed quantization schemes designed under the Gaussian assumption. For rates much below 1 bit/pixel the relative improvement is negligible.",
author = "Modestino, {James W.} and Nariman Farvardin and Ogrinc, {Michael A.}",
year = "1985",
month = "3",
day = "1",
language = "English",
volume = "COM-33",
pages = "210--217",
journal = "IEEE Transactions on Communications",
issn = "0096-1965",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "3",

}

TY - JOUR

T1 - PERFORMANCE OF BLOCK COSINE IMAGE CODING WITH ADAPTIVE QUANTIZATION.

AU - Modestino, James W.

AU - Farvardin, Nariman

AU - Ogrinc, Michael A.

PY - 1985/3/1

Y1 - 1985/3/1

N2 - Quantizers for block transform image coding systems are typically designed under the assumption of Gaussian statistics for the transform coefficients. While convincing arguments can be provided in support of this approach, empirical evidence is presented demonstrating that, except possibly for the dc term, wide departures from Gaussian behavior can be expected for real-world imagery at typical block sizes. In this paper we describe the performance of a block cosine image coding system with an adaptive quantizer matched to the statistics of the transform coefficients. The adaptive quantizer is based upon a recently developed algorithm which employs a training sequence in the design procedure. At encoding rates of approximately 1 bit/pixel and above, this approach results in significant improvement in reconstructed image quality compared to fixed quantization schemes designed under the Gaussian assumption. For rates much below 1 bit/pixel the relative improvement is negligible.

AB - Quantizers for block transform image coding systems are typically designed under the assumption of Gaussian statistics for the transform coefficients. While convincing arguments can be provided in support of this approach, empirical evidence is presented demonstrating that, except possibly for the dc term, wide departures from Gaussian behavior can be expected for real-world imagery at typical block sizes. In this paper we describe the performance of a block cosine image coding system with an adaptive quantizer matched to the statistics of the transform coefficients. The adaptive quantizer is based upon a recently developed algorithm which employs a training sequence in the design procedure. At encoding rates of approximately 1 bit/pixel and above, this approach results in significant improvement in reconstructed image quality compared to fixed quantization schemes designed under the Gaussian assumption. For rates much below 1 bit/pixel the relative improvement is negligible.

UR - http://www.scopus.com/inward/record.url?scp=0022028250&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0022028250&partnerID=8YFLogxK

M3 - Article

VL - COM-33

SP - 210

EP - 217

JO - IEEE Transactions on Communications

JF - IEEE Transactions on Communications

SN - 0096-1965

IS - 3

ER -