Exploiting image indexing techniques in DCT domain

C. W. Ngo; T. C. Pong; R. T. Chin

doi:10.1007/BFb0016499

Exploiting image indexing techniques in DCT domain

C. W. Ngo, T. C. Pong, R. T. Chin

Department of Computer Science and Engineering

Research output: Chapter in Book/Conference Proceeding/Report › Conference Paper published in a book › peer-review

6 Citations (Scopus)

Abstract

This paper is concerned with the indexing and retrieval of images based on features extracted directly from JPEG discrete cosine transform (DCT) domain. We examine possible ways of manipulating DCT coefficients by standard image analysis approaches to describe image shape, texture, and color. Through Mandala transformation, our approach groups a subset of DCT coefficients to form ten blocks. Each block represents a particular frequency of the original image. We use two blocks to model rough object shape; nine blocks to describe subband properties; and one block to compute color distribution. As a result, the amount of data used for processing and analysis is reduced significantly. This can lead to simple yet efficient ways, of indexing and retrieval in a large scale image database. Experimental results show that it only takes approximately 6ms to index shape features, 5ms to index texture features, and 8ms to index color features from an image the size of 128 x 128 on a Sun Sparc Ultra-1 machine.

Original language	English
Title of host publication	Multimedia Information Analysis and Retrieval - IAPR International Workshop, MINAR 1998, Proceedings
Editors	Horace H.S. Ip, Arnold W.M. Smeulders
Publisher	Springer Verlag
Pages	195-206
Number of pages	12
ISBN (Print)	3540648267, 9783540648260
DOIs	https://doi.org/10.1007/BFb0016499
Publication status	Published - 1998
Event	IAPR International Workshop on Multimedia Information Analysis and Retrieval, MINAR 1998 - Hong Kong, China Duration: 13 Aug 1998 → 14 Aug 1998

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	1464
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	IAPR International Workshop on Multimedia Information Analysis and Retrieval, MINAR 1998
Country/Territory	China
City	Hong Kong
Period	13/08/98 → 14/08/98

Bibliographical note

Publisher Copyright:
© Springer-Verlag Berlin Heidelberg 1998.

Access to Document

10.1007/BFb0016499

Cite this

Ngo, C. W., Pong, T. C., & Chin, R. T. (1998). Exploiting image indexing techniques in DCT domain. In H. H. S. Ip, & A. W. M. Smeulders (Eds.), Multimedia Information Analysis and Retrieval - IAPR International Workshop, MINAR 1998, Proceedings (pp. 195-206). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1464). Springer Verlag. https://doi.org/10.1007/BFb0016499

Ngo, C. W. ; Pong, T. C. ; Chin, R. T. / Exploiting image indexing techniques in DCT domain. Multimedia Information Analysis and Retrieval - IAPR International Workshop, MINAR 1998, Proceedings. editor / Horace H.S. Ip ; Arnold W.M. Smeulders. Springer Verlag, 1998. pp. 195-206 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{f27999efc7c94aa0bfc6cfc27655b6bb,

title = "Exploiting image indexing techniques in DCT domain",

abstract = "This paper is concerned with the indexing and retrieval of images based on features extracted directly from JPEG discrete cosine transform (DCT) domain. We examine possible ways of manipulating DCT coefficients by standard image analysis approaches to describe image shape, texture, and color. Through Mandala transformation, our approach groups a subset of DCT coefficients to form ten blocks. Each block represents a particular frequency of the original image. We use two blocks to model rough object shape; nine blocks to describe subband properties; and one block to compute color distribution. As a result, the amount of data used for processing and analysis is reduced significantly. This can lead to simple yet efficient ways, of indexing and retrieval in a large scale image database. Experimental results show that it only takes approximately 6ms to index shape features, 5ms to index texture features, and 8ms to index color features from an image the size of 128 x 128 on a Sun Sparc Ultra-1 machine.",

author = "Ngo, \{C. W.\} and Pong, \{T. C.\} and Chin, \{R. T.\}",

note = "Publisher Copyright: {\textcopyright} Springer-Verlag Berlin Heidelberg 1998.; IAPR International Workshop on Multimedia Information Analysis and Retrieval, MINAR 1998 ; Conference date: 13-08-1998 Through 14-08-1998",

year = "1998",

doi = "10.1007/BFb0016499",

language = "English",

isbn = "3540648267",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "195--206",

editor = "Ip, \{Horace H.S.\} and Smeulders, \{Arnold W.M.\}",

booktitle = "Multimedia Information Analysis and Retrieval - IAPR International Workshop, MINAR 1998, Proceedings",

}

Ngo, CW, Pong, TC & Chin, RT 1998, Exploiting image indexing techniques in DCT domain. in HHS Ip & AWM Smeulders (eds), Multimedia Information Analysis and Retrieval - IAPR International Workshop, MINAR 1998, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 1464, Springer Verlag, pp. 195-206, IAPR International Workshop on Multimedia Information Analysis and Retrieval, MINAR 1998, Hong Kong, China, 13/08/98. https://doi.org/10.1007/BFb0016499

Exploiting image indexing techniques in DCT domain. / Ngo, C. W.; Pong, T. C.; Chin, R. T.
Multimedia Information Analysis and Retrieval - IAPR International Workshop, MINAR 1998, Proceedings. ed. / Horace H.S. Ip; Arnold W.M. Smeulders. Springer Verlag, 1998. p. 195-206 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1464).

Research output: Chapter in Book/Conference Proceeding/Report › Conference Paper published in a book › peer-review

TY - GEN

T1 - Exploiting image indexing techniques in DCT domain

AU - Ngo, C. W.

AU - Pong, T. C.

AU - Chin, R. T.

N1 - Publisher Copyright: © Springer-Verlag Berlin Heidelberg 1998.

PY - 1998

Y1 - 1998

N2 - This paper is concerned with the indexing and retrieval of images based on features extracted directly from JPEG discrete cosine transform (DCT) domain. We examine possible ways of manipulating DCT coefficients by standard image analysis approaches to describe image shape, texture, and color. Through Mandala transformation, our approach groups a subset of DCT coefficients to form ten blocks. Each block represents a particular frequency of the original image. We use two blocks to model rough object shape; nine blocks to describe subband properties; and one block to compute color distribution. As a result, the amount of data used for processing and analysis is reduced significantly. This can lead to simple yet efficient ways, of indexing and retrieval in a large scale image database. Experimental results show that it only takes approximately 6ms to index shape features, 5ms to index texture features, and 8ms to index color features from an image the size of 128 x 128 on a Sun Sparc Ultra-1 machine.

AB - This paper is concerned with the indexing and retrieval of images based on features extracted directly from JPEG discrete cosine transform (DCT) domain. We examine possible ways of manipulating DCT coefficients by standard image analysis approaches to describe image shape, texture, and color. Through Mandala transformation, our approach groups a subset of DCT coefficients to form ten blocks. Each block represents a particular frequency of the original image. We use two blocks to model rough object shape; nine blocks to describe subband properties; and one block to compute color distribution. As a result, the amount of data used for processing and analysis is reduced significantly. This can lead to simple yet efficient ways, of indexing and retrieval in a large scale image database. Experimental results show that it only takes approximately 6ms to index shape features, 5ms to index texture features, and 8ms to index color features from an image the size of 128 x 128 on a Sun Sparc Ultra-1 machine.

UR - https://www.scopus.com/pages/publications/84886239894

U2 - 10.1007/BFb0016499

DO - 10.1007/BFb0016499

M3 - Conference Paper published in a book

AN - SCOPUS:84886239894

SN - 3540648267

SN - 9783540648260

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 195

EP - 206

BT - Multimedia Information Analysis and Retrieval - IAPR International Workshop, MINAR 1998, Proceedings

A2 - Ip, Horace H.S.

A2 - Smeulders, Arnold W.M.

PB - Springer Verlag

T2 - IAPR International Workshop on Multimedia Information Analysis and Retrieval, MINAR 1998

Y2 - 13 August 1998 through 14 August 1998

ER -

Ngo CW, Pong TC, Chin RT. Exploiting image indexing techniques in DCT domain. In Ip HHS, Smeulders AWM, editors, Multimedia Information Analysis and Retrieval - IAPR International Workshop, MINAR 1998, Proceedings. Springer Verlag. 1998. p. 195-206. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/BFb0016499

Exploiting image indexing techniques in DCT domain

Abstract

Publication series

Conference

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this