Algorithms and architecture for image adaptive vector quantization

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

In this paper, we present two algorithms for vector quantization of images and an architecture to implement these algorithms. In vector quantization (VQ), the image vectors are usually coded with an “universal” codebook, however, for a given image, only a subset of the codewords in the universal codebook may be needed. This means that effectively a smaller label size can be employed at the expense of a small overhead information to indicate to the receiver the codewords used. Simulation results demonstrate the superior coding performance of this technique. VQ using an universal codebook (VQUC) is computationally less demanding but its performance is poor for images outside the training sequence. Image adaptive techniques, where new codebooks are generated, for each input image (VQIAC) can improve the performance but at the cost of increased computational complexity. A technique which combines the advantages of VQUC and VQIAC is presented in this paper. Simulation results demonstrate that the technique gives a coding performance close to that obtained with image adaptive VQ at a substantially reduced computational complextiy. A systolic array architecture to implement the algorithms in real-time is also presented. The regular and iterable structure makes possible the VLSI implementation of the architecture.

Original languageEnglish (US)
Pages (from-to)336-344
Number of pages9
JournalProceedings of SPIE - The International Society for Optical Engineering
Volume1001
DOIs
StatePublished - Oct 25 1988
Externally publishedYes

Fingerprint

vector quantization
Vector Quantization
Vector quantization
Codebook
Systolic arrays
coding
Coding
Labels
Computational complexity
systolic arrays
Systolic Array
Adaptive Techniques
very large scale integration
Architecture
Demonstrate
set theory
Computational Complexity
Simulation
education
Receiver

ASJC Scopus subject areas

  • Applied Mathematics
  • Electronic, Optical and Magnetic Materials
  • Condensed Matter Physics
  • Electrical and Electronic Engineering
  • Computer Science Applications

Cite this

Algorithms and architecture for image adaptive vector quantization. / Panchanathan, Sethuraman; Goldberg, M.

In: Proceedings of SPIE - The International Society for Optical Engineering, Vol. 1001, 25.10.1988, p. 336-344.

Research output: Contribution to journalArticle

@article{d148ed35a65c4d66a3207eb89120584f,
title = "Algorithms and architecture for image adaptive vector quantization",
abstract = "In this paper, we present two algorithms for vector quantization of images and an architecture to implement these algorithms. In vector quantization (VQ), the image vectors are usually coded with an “universal” codebook, however, for a given image, only a subset of the codewords in the universal codebook may be needed. This means that effectively a smaller label size can be employed at the expense of a small overhead information to indicate to the receiver the codewords used. Simulation results demonstrate the superior coding performance of this technique. VQ using an universal codebook (VQUC) is computationally less demanding but its performance is poor for images outside the training sequence. Image adaptive techniques, where new codebooks are generated, for each input image (VQIAC) can improve the performance but at the cost of increased computational complexity. A technique which combines the advantages of VQUC and VQIAC is presented in this paper. Simulation results demonstrate that the technique gives a coding performance close to that obtained with image adaptive VQ at a substantially reduced computational complextiy. A systolic array architecture to implement the algorithms in real-time is also presented. The regular and iterable structure makes possible the VLSI implementation of the architecture.",
author = "Sethuraman Panchanathan and M. Goldberg",
year = "1988",
month = "10",
day = "25",
doi = "10.1117/12.968971",
language = "English (US)",
volume = "1001",
pages = "336--344",
journal = "Proceedings of SPIE - The International Society for Optical Engineering",
issn = "0277-786X",
publisher = "SPIE",

}

TY - JOUR

T1 - Algorithms and architecture for image adaptive vector quantization

AU - Panchanathan, Sethuraman

AU - Goldberg, M.

PY - 1988/10/25

Y1 - 1988/10/25

N2 - In this paper, we present two algorithms for vector quantization of images and an architecture to implement these algorithms. In vector quantization (VQ), the image vectors are usually coded with an “universal” codebook, however, for a given image, only a subset of the codewords in the universal codebook may be needed. This means that effectively a smaller label size can be employed at the expense of a small overhead information to indicate to the receiver the codewords used. Simulation results demonstrate the superior coding performance of this technique. VQ using an universal codebook (VQUC) is computationally less demanding but its performance is poor for images outside the training sequence. Image adaptive techniques, where new codebooks are generated, for each input image (VQIAC) can improve the performance but at the cost of increased computational complexity. A technique which combines the advantages of VQUC and VQIAC is presented in this paper. Simulation results demonstrate that the technique gives a coding performance close to that obtained with image adaptive VQ at a substantially reduced computational complextiy. A systolic array architecture to implement the algorithms in real-time is also presented. The regular and iterable structure makes possible the VLSI implementation of the architecture.

AB - In this paper, we present two algorithms for vector quantization of images and an architecture to implement these algorithms. In vector quantization (VQ), the image vectors are usually coded with an “universal” codebook, however, for a given image, only a subset of the codewords in the universal codebook may be needed. This means that effectively a smaller label size can be employed at the expense of a small overhead information to indicate to the receiver the codewords used. Simulation results demonstrate the superior coding performance of this technique. VQ using an universal codebook (VQUC) is computationally less demanding but its performance is poor for images outside the training sequence. Image adaptive techniques, where new codebooks are generated, for each input image (VQIAC) can improve the performance but at the cost of increased computational complexity. A technique which combines the advantages of VQUC and VQIAC is presented in this paper. Simulation results demonstrate that the technique gives a coding performance close to that obtained with image adaptive VQ at a substantially reduced computational complextiy. A systolic array architecture to implement the algorithms in real-time is also presented. The regular and iterable structure makes possible the VLSI implementation of the architecture.

UR - http://www.scopus.com/inward/record.url?scp=84940925905&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84940925905&partnerID=8YFLogxK

U2 - 10.1117/12.968971

DO - 10.1117/12.968971

M3 - Article

AN - SCOPUS:84940925905

VL - 1001

SP - 336

EP - 344

JO - Proceedings of SPIE - The International Society for Optical Engineering

JF - Proceedings of SPIE - The International Society for Optical Engineering

SN - 0277-786X

ER -