Content-based image indexing and searching using Daubechies' wavelets

James Wang, Gio Wiederhold, Oscar Firschein, Sha Xin Wei

Research output: Contribution to journalArticle

282 Citations (Scopus)

Abstract

This paper describes WBIIS (Wavelet-Based Image Indexing and Searching), a new image indexing and retrieval algorithm with partial sketch image searching capability for large image databases. The algorithm characterizes the color variations over the spatial extent of the image in a manner that provides semantically meaningful image comparisons. The indexing algorithm applies a Daubechies' wavelet transform for each of the three opponent color components. The wavelet coefficients in the lowest few frequency bands, and their variances, are stored as feature vectors. To speed up retrieval, a two-step procedure is used that first does a crude selection based on the variances, and then refines the search by performing a feature vector match between the selected images and the query. For better accuracy in searching, two-level multiresolution matching may also be used. Masks are used for partial-sketch queries. This technique performs much better in capturing coherence of image, object granularity, local color/texture, and bias avoidance than traditional color layout algorithms. WBIIS is much faster and more accurate than traditional algorithms. When tested on a database of more than 10 000 general-purpose images, the best 100 matches were found in 3.3 seconds.

Original languageEnglish (US)
Pages (from-to)311-328
Number of pages18
JournalInternational Journal on Digital Libraries
Volume1
Issue number4
DOIs
StatePublished - Dec 1 1997

Fingerprint

indexing
layout
trend

All Science Journal Classification (ASJC) codes

  • Library and Information Sciences

Cite this

Wang, James ; Wiederhold, Gio ; Firschein, Oscar ; Wei, Sha Xin. / Content-based image indexing and searching using Daubechies' wavelets. In: International Journal on Digital Libraries. 1997 ; Vol. 1, No. 4. pp. 311-328.
@article{e17bcbf3737c4985a82264cd93f524d6,
title = "Content-based image indexing and searching using Daubechies' wavelets",
abstract = "This paper describes WBIIS (Wavelet-Based Image Indexing and Searching), a new image indexing and retrieval algorithm with partial sketch image searching capability for large image databases. The algorithm characterizes the color variations over the spatial extent of the image in a manner that provides semantically meaningful image comparisons. The indexing algorithm applies a Daubechies' wavelet transform for each of the three opponent color components. The wavelet coefficients in the lowest few frequency bands, and their variances, are stored as feature vectors. To speed up retrieval, a two-step procedure is used that first does a crude selection based on the variances, and then refines the search by performing a feature vector match between the selected images and the query. For better accuracy in searching, two-level multiresolution matching may also be used. Masks are used for partial-sketch queries. This technique performs much better in capturing coherence of image, object granularity, local color/texture, and bias avoidance than traditional color layout algorithms. WBIIS is much faster and more accurate than traditional algorithms. When tested on a database of more than 10 000 general-purpose images, the best 100 matches were found in 3.3 seconds.",
author = "James Wang and Gio Wiederhold and Oscar Firschein and Wei, {Sha Xin}",
year = "1997",
month = "12",
day = "1",
doi = "10.1007/s007990050026",
language = "English (US)",
volume = "1",
pages = "311--328",
journal = "International Journal on Digital Libraries",
issn = "1432-5012",
publisher = "Springer Verlag",
number = "4",

}

Content-based image indexing and searching using Daubechies' wavelets. / Wang, James; Wiederhold, Gio; Firschein, Oscar; Wei, Sha Xin.

In: International Journal on Digital Libraries, Vol. 1, No. 4, 01.12.1997, p. 311-328.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Content-based image indexing and searching using Daubechies' wavelets

AU - Wang, James

AU - Wiederhold, Gio

AU - Firschein, Oscar

AU - Wei, Sha Xin

PY - 1997/12/1

Y1 - 1997/12/1

N2 - This paper describes WBIIS (Wavelet-Based Image Indexing and Searching), a new image indexing and retrieval algorithm with partial sketch image searching capability for large image databases. The algorithm characterizes the color variations over the spatial extent of the image in a manner that provides semantically meaningful image comparisons. The indexing algorithm applies a Daubechies' wavelet transform for each of the three opponent color components. The wavelet coefficients in the lowest few frequency bands, and their variances, are stored as feature vectors. To speed up retrieval, a two-step procedure is used that first does a crude selection based on the variances, and then refines the search by performing a feature vector match between the selected images and the query. For better accuracy in searching, two-level multiresolution matching may also be used. Masks are used for partial-sketch queries. This technique performs much better in capturing coherence of image, object granularity, local color/texture, and bias avoidance than traditional color layout algorithms. WBIIS is much faster and more accurate than traditional algorithms. When tested on a database of more than 10 000 general-purpose images, the best 100 matches were found in 3.3 seconds.

AB - This paper describes WBIIS (Wavelet-Based Image Indexing and Searching), a new image indexing and retrieval algorithm with partial sketch image searching capability for large image databases. The algorithm characterizes the color variations over the spatial extent of the image in a manner that provides semantically meaningful image comparisons. The indexing algorithm applies a Daubechies' wavelet transform for each of the three opponent color components. The wavelet coefficients in the lowest few frequency bands, and their variances, are stored as feature vectors. To speed up retrieval, a two-step procedure is used that first does a crude selection based on the variances, and then refines the search by performing a feature vector match between the selected images and the query. For better accuracy in searching, two-level multiresolution matching may also be used. Masks are used for partial-sketch queries. This technique performs much better in capturing coherence of image, object granularity, local color/texture, and bias avoidance than traditional color layout algorithms. WBIIS is much faster and more accurate than traditional algorithms. When tested on a database of more than 10 000 general-purpose images, the best 100 matches were found in 3.3 seconds.

UR - http://www.scopus.com/inward/record.url?scp=84867521696&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84867521696&partnerID=8YFLogxK

U2 - 10.1007/s007990050026

DO - 10.1007/s007990050026

M3 - Article

VL - 1

SP - 311

EP - 328

JO - International Journal on Digital Libraries

JF - International Journal on Digital Libraries

SN - 1432-5012

IS - 4

ER -