Scan predictive vector quantization of multispectral images

Nasir D. Memon, Khalid Sayood

Research output: Contribution to journalArticle

Abstract

Conventional vector quantization (VQ)-based techniques partition an image into nonoverlapping blocks that are then raster scanned and quantized. Image blocks that contain an edge result in high-frequency vectors. The coarse representation of such vectors leads to visually annoying degradations in the reconstructed image. We present a solution to the edge-degradation problem based on some earlier work on scan models. Our approach reduces the number of vectors with abrupt intensity variations by using an appropriate scan to partition an image into vectors. We show how our techniques can be used to enhance the performance of VQ of multispectral data sets. Comparisons with standard techniques are presented and shown to give substantial improvements.

Original languageEnglish (US)
Pages (from-to)330-337
Number of pages8
JournalIEEE Transactions on Image Processing
Volume5
Issue number2
DOIs
StatePublished - 1996

Fingerprint

Multispectral Images
Vector Quantization
Vector quantization
Degradation
Partition

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Computer Graphics and Computer-Aided Design
  • Software
  • Theoretical Computer Science
  • Computational Theory and Mathematics
  • Computer Vision and Pattern Recognition

Cite this

Scan predictive vector quantization of multispectral images. / Memon, Nasir D.; Sayood, Khalid.

In: IEEE Transactions on Image Processing, Vol. 5, No. 2, 1996, p. 330-337.

Research output: Contribution to journalArticle

@article{b9cff3c68932401c91a9a4c82aab53e1,
title = "Scan predictive vector quantization of multispectral images",
abstract = "Conventional vector quantization (VQ)-based techniques partition an image into nonoverlapping blocks that are then raster scanned and quantized. Image blocks that contain an edge result in high-frequency vectors. The coarse representation of such vectors leads to visually annoying degradations in the reconstructed image. We present a solution to the edge-degradation problem based on some earlier work on scan models. Our approach reduces the number of vectors with abrupt intensity variations by using an appropriate scan to partition an image into vectors. We show how our techniques can be used to enhance the performance of VQ of multispectral data sets. Comparisons with standard techniques are presented and shown to give substantial improvements.",
author = "Memon, {Nasir D.} and Khalid Sayood",
year = "1996",
doi = "10.1109/83.480768",
language = "English (US)",
volume = "5",
pages = "330--337",
journal = "IEEE Transactions on Image Processing",
issn = "1057-7149",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "2",

}

TY - JOUR

T1 - Scan predictive vector quantization of multispectral images

AU - Memon, Nasir D.

AU - Sayood, Khalid

PY - 1996

Y1 - 1996

N2 - Conventional vector quantization (VQ)-based techniques partition an image into nonoverlapping blocks that are then raster scanned and quantized. Image blocks that contain an edge result in high-frequency vectors. The coarse representation of such vectors leads to visually annoying degradations in the reconstructed image. We present a solution to the edge-degradation problem based on some earlier work on scan models. Our approach reduces the number of vectors with abrupt intensity variations by using an appropriate scan to partition an image into vectors. We show how our techniques can be used to enhance the performance of VQ of multispectral data sets. Comparisons with standard techniques are presented and shown to give substantial improvements.

AB - Conventional vector quantization (VQ)-based techniques partition an image into nonoverlapping blocks that are then raster scanned and quantized. Image blocks that contain an edge result in high-frequency vectors. The coarse representation of such vectors leads to visually annoying degradations in the reconstructed image. We present a solution to the edge-degradation problem based on some earlier work on scan models. Our approach reduces the number of vectors with abrupt intensity variations by using an appropriate scan to partition an image into vectors. We show how our techniques can be used to enhance the performance of VQ of multispectral data sets. Comparisons with standard techniques are presented and shown to give substantial improvements.

UR - http://www.scopus.com/inward/record.url?scp=0030081228&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0030081228&partnerID=8YFLogxK

U2 - 10.1109/83.480768

DO - 10.1109/83.480768

M3 - Article

C2 - 18285116

AN - SCOPUS:0030081228

VL - 5

SP - 330

EP - 337

JO - IEEE Transactions on Image Processing

JF - IEEE Transactions on Image Processing

SN - 1057-7149

IS - 2

ER -