Dynamic windowed codebook search algorithm in vector quantization

Yih Chuan Lin, Shen-Chuan Tai

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

We present an efficient codebook search algorithm in a vector quantization (VQ)-based system that searches only through a window of consecutive codewords on the preordered codebook, rather than all codewords in the codebook, and achieves full-search-equivalent performance. The ordering technique used to rearrange the positions of codewords in the codebook utilizes the scattering orientation of codewords in the k-dimensional space or the compactness property of signal energy on the transform domain. The procedure for determining the search window on the ordered codebook relies on the geometric relations among input vector and codewords to eliminate those codewords that are certain not to be the closest codeword to the input vector. Finally, the closest codeword to an input vector is selected from the determined search window by an improved partial distortion search algorithm. In comparison with other existing fast algorithms, the proposed algorithm requires the least number of multiplications and the least total number of distortion measurements.

Original languageEnglish
Article number33105
Pages (from-to)2921-2929
Number of pages9
JournalOptical Engineering
Volume35
Issue number10
DOIs
Publication statusPublished - 1996 Jan 1

Fingerprint

vector quantization
Vector quantization
void ratio
multiplication
Scattering
scattering
energy

All Science Journal Classification (ASJC) codes

  • Atomic and Molecular Physics, and Optics
  • Engineering(all)

Cite this

@article{2d8be449dde24cb7830420ba79dc1fe5,
title = "Dynamic windowed codebook search algorithm in vector quantization",
abstract = "We present an efficient codebook search algorithm in a vector quantization (VQ)-based system that searches only through a window of consecutive codewords on the preordered codebook, rather than all codewords in the codebook, and achieves full-search-equivalent performance. The ordering technique used to rearrange the positions of codewords in the codebook utilizes the scattering orientation of codewords in the k-dimensional space or the compactness property of signal energy on the transform domain. The procedure for determining the search window on the ordered codebook relies on the geometric relations among input vector and codewords to eliminate those codewords that are certain not to be the closest codeword to the input vector. Finally, the closest codeword to an input vector is selected from the determined search window by an improved partial distortion search algorithm. In comparison with other existing fast algorithms, the proposed algorithm requires the least number of multiplications and the least total number of distortion measurements.",
author = "Lin, {Yih Chuan} and Shen-Chuan Tai",
year = "1996",
month = "1",
day = "1",
doi = "10.1117/1.600976",
language = "English",
volume = "35",
pages = "2921--2929",
journal = "Optical Engineering",
issn = "0091-3286",
publisher = "SPIE",
number = "10",

}

Dynamic windowed codebook search algorithm in vector quantization. / Lin, Yih Chuan; Tai, Shen-Chuan.

In: Optical Engineering, Vol. 35, No. 10, 33105, 01.01.1996, p. 2921-2929.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Dynamic windowed codebook search algorithm in vector quantization

AU - Lin, Yih Chuan

AU - Tai, Shen-Chuan

PY - 1996/1/1

Y1 - 1996/1/1

N2 - We present an efficient codebook search algorithm in a vector quantization (VQ)-based system that searches only through a window of consecutive codewords on the preordered codebook, rather than all codewords in the codebook, and achieves full-search-equivalent performance. The ordering technique used to rearrange the positions of codewords in the codebook utilizes the scattering orientation of codewords in the k-dimensional space or the compactness property of signal energy on the transform domain. The procedure for determining the search window on the ordered codebook relies on the geometric relations among input vector and codewords to eliminate those codewords that are certain not to be the closest codeword to the input vector. Finally, the closest codeword to an input vector is selected from the determined search window by an improved partial distortion search algorithm. In comparison with other existing fast algorithms, the proposed algorithm requires the least number of multiplications and the least total number of distortion measurements.

AB - We present an efficient codebook search algorithm in a vector quantization (VQ)-based system that searches only through a window of consecutive codewords on the preordered codebook, rather than all codewords in the codebook, and achieves full-search-equivalent performance. The ordering technique used to rearrange the positions of codewords in the codebook utilizes the scattering orientation of codewords in the k-dimensional space or the compactness property of signal energy on the transform domain. The procedure for determining the search window on the ordered codebook relies on the geometric relations among input vector and codewords to eliminate those codewords that are certain not to be the closest codeword to the input vector. Finally, the closest codeword to an input vector is selected from the determined search window by an improved partial distortion search algorithm. In comparison with other existing fast algorithms, the proposed algorithm requires the least number of multiplications and the least total number of distortion measurements.

UR - http://www.scopus.com/inward/record.url?scp=0000118609&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0000118609&partnerID=8YFLogxK

U2 - 10.1117/1.600976

DO - 10.1117/1.600976

M3 - Article

AN - SCOPUS:0000118609

VL - 35

SP - 2921

EP - 2929

JO - Optical Engineering

JF - Optical Engineering

SN - 0091-3286

IS - 10

M1 - 33105

ER -