Coding Gain in Non-Paraunitary Subband Coding Systems

S. A. Asghar BEHESHTI SHIRAZI; Yoshitaka MORIKAWA; Hiroshi HAMADA

Coding Gain in Non-Paraunitary Subband Coding Systems

S. A. Asghar BEHESHTI SHIRAZI, Yoshitaka MORIKAWA, Hiroshi HAMADA

Full Text Views

0

Share
Cite this

Summary :

This work addresses the problems of bit allocation and coding gain in subband coding system with non-paraunitary filter banks. Since energy conservation does not hold in non-paraunitary filter banks, the model to be adopted for quantizers is important to evaluate the output distortion introduced by subband signal quantization. To evaluate the overall distortion we start with adopting the gain plus additive noise model for quantizers, which is more reliable than the additive noise model. With this model, the expression for overall reconstruction error variance becomes so complicated that the problem of optimum bit allocation, as required for evaluation of the coding gain, must be numerically solved. So, we propose an approximation method in which we neglect the terms due to correlation among quantization errors in calculating the bit allocation but take them into consideration in evaluating the coding gain, assuming sufficiently high bitrate coding. Application of this approximation method to the SSKF subband coding systems with AR (1) input source shows that the method is very accurate even at low bit rate coding (1 bit/sample).

Publication: IEICE TRANSACTIONS on Fundamentals Vol.E79-A No.2 pp.233-241

Publication Date: 1996/02/25

Publicized

Online ISSN

DOI

Type of Manuscript: PAPER

Category: Information Theory and Coding Theory

Cite this

Copy

S. A. Asghar BEHESHTI SHIRAZI, Yoshitaka MORIKAWA, Hiroshi HAMADA, "Coding Gain in Non-Paraunitary Subband Coding Systems" in IEICE TRANSACTIONS on Fundamentals, vol. E79-A, no. 2, pp. 233-241, February 1996, doi: .
Abstract: This work addresses the problems of bit allocation and coding gain in subband coding system with non-paraunitary filter banks. Since energy conservation does not hold in non-paraunitary filter banks, the model to be adopted for quantizers is important to evaluate the output distortion introduced by subband signal quantization. To evaluate the overall distortion we start with adopting the gain plus additive noise model for quantizers, which is more reliable than the additive noise model. With this model, the expression for overall reconstruction error variance becomes so complicated that the problem of optimum bit allocation, as required for evaluation of the coding gain, must be numerically solved. So, we propose an approximation method in which we neglect the terms due to correlation among quantization errors in calculating the bit allocation but take them into consideration in evaluating the coding gain, assuming sufficiently high bitrate coding. Application of this approximation method to the SSKF subband coding systems with AR (1) input source shows that the method is very accurate even at low bit rate coding (1 bit/sample).
URL: https://globals.ieice.org/en_transactions/fundamentals/10.1587/e79-a_2_233/_p

Copy

@ARTICLE{e79-a_2_233,
author={S. A. Asghar BEHESHTI SHIRAZI, Yoshitaka MORIKAWA, Hiroshi HAMADA, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={Coding Gain in Non-Paraunitary Subband Coding Systems},
year={1996},
volume={E79-A},
number={2},
pages={233-241},
abstract={This work addresses the problems of bit allocation and coding gain in subband coding system with non-paraunitary filter banks. Since energy conservation does not hold in non-paraunitary filter banks, the model to be adopted for quantizers is important to evaluate the output distortion introduced by subband signal quantization. To evaluate the overall distortion we start with adopting the gain plus additive noise model for quantizers, which is more reliable than the additive noise model. With this model, the expression for overall reconstruction error variance becomes so complicated that the problem of optimum bit allocation, as required for evaluation of the coding gain, must be numerically solved. So, we propose an approximation method in which we neglect the terms due to correlation among quantization errors in calculating the bit allocation but take them into consideration in evaluating the coding gain, assuming sufficiently high bitrate coding. Application of this approximation method to the SSKF subband coding systems with AR (1) input source shows that the method is very accurate even at low bit rate coding (1 bit/sample).},
keywords={},
doi={},
ISSN={},
month={February},}

Copy

TY - JOUR
TI - Coding Gain in Non-Paraunitary Subband Coding Systems
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 233
EP - 241
AU - S. A. Asghar BEHESHTI SHIRAZI
AU - Yoshitaka MORIKAWA
AU - Hiroshi HAMADA
PY - 1996
DO -
JO - IEICE TRANSACTIONS on Fundamentals
SN -
VL - E79-A
IS - 2
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - February 1996
AB - This work addresses the problems of bit allocation and coding gain in subband coding system with non-paraunitary filter banks. Since energy conservation does not hold in non-paraunitary filter banks, the model to be adopted for quantizers is important to evaluate the output distortion introduced by subband signal quantization. To evaluate the overall distortion we start with adopting the gain plus additive noise model for quantizers, which is more reliable than the additive noise model. With this model, the expression for overall reconstruction error variance becomes so complicated that the problem of optimum bit allocation, as required for evaluation of the coding gain, must be numerically solved. So, we propose an approximation method in which we neglect the terms due to correlation among quantization errors in calculating the bit allocation but take them into consideration in evaluating the coding gain, assuming sufficiently high bitrate coding. Application of this approximation method to the SSKF subband coding systems with AR (1) input source shows that the method is very accurate even at low bit rate coding (1 bit/sample).
ER -