FIRM: Fuzzily integrated region matching for content-based image retrieval

Y. Chen; J. Z. Wang; J. Li

FIRM: Fuzzily integrated region matching for content-based image retrieval

Research output: Contribution to conference › Paper › peer-review

Abstract

We propose FIRM (Fuzzily Integrated Region Matching), an efficient and robust similarity measure for region-based image retrieval. Each image in our retrieval system is represented by a set of regions that are characterized by fuzzy sets. The FIRM measure, representing the overall similarity between two images, is defined as the similarity between two families of fuzzy sets. Compared with similarity measures based on individual regions and on all regions with crisp feature representations, our approach greatly reduces the influence of inaccurate segmentation. Experimental results based on a database of about 200,000 general-purpose images demonstrate improved accuracy, robustness, and high speed.

Original language	English (US)
Pages	543-545
Number of pages	3
State	Published - Jan 1 2001
Event	-ACM Multimedia 2001 Workshops- 2001 Multimedia Conference - Ottawa, Ont., Canada Duration: Sep 30 2001 → Oct 5 2001

Other

Other	-ACM Multimedia 2001 Workshops- 2001 Multimedia Conference
Country	Canada
City	Ottawa, Ont.
Period	9/30/01 → 10/5/01

All Science Journal Classification (ASJC) codes

Computer Science(all)

Fingerprint Dive into the research topics of 'FIRM: Fuzzily integrated region matching for content-based image retrieval'. Together they form a unique fingerprint.

Cite this

@conference{52b60dd0e09543afae6288e489490d54,

title = "FIRM: Fuzzily integrated region matching for content-based image retrieval",

abstract = "We propose FIRM (Fuzzily Integrated Region Matching), an efficient and robust similarity measure for region-based image retrieval. Each image in our retrieval system is represented by a set of regions that are characterized by fuzzy sets. The FIRM measure, representing the overall similarity between two images, is defined as the similarity between two families of fuzzy sets. Compared with similarity measures based on individual regions and on all regions with crisp feature representations, our approach greatly reduces the influence of inaccurate segmentation. Experimental results based on a database of about 200,000 general-purpose images demonstrate improved accuracy, robustness, and high speed.",

author = "Y. Chen and Wang, {J. Z.} and J. Li",

year = "2001",

month = jan,

day = "1",

language = "English (US)",

pages = "543--545",

note = "-ACM Multimedia 2001 Workshops- 2001 Multimedia Conference ; Conference date: 30-09-2001 Through 05-10-2001",

}

TY - CONF

T1 - FIRM

T2 - -ACM Multimedia 2001 Workshops- 2001 Multimedia Conference

AU - Chen, Y.

AU - Wang, J. Z.

AU - Li, J.

PY - 2001/1/1

Y1 - 2001/1/1

N2 - We propose FIRM (Fuzzily Integrated Region Matching), an efficient and robust similarity measure for region-based image retrieval. Each image in our retrieval system is represented by a set of regions that are characterized by fuzzy sets. The FIRM measure, representing the overall similarity between two images, is defined as the similarity between two families of fuzzy sets. Compared with similarity measures based on individual regions and on all regions with crisp feature representations, our approach greatly reduces the influence of inaccurate segmentation. Experimental results based on a database of about 200,000 general-purpose images demonstrate improved accuracy, robustness, and high speed.

AB - We propose FIRM (Fuzzily Integrated Region Matching), an efficient and robust similarity measure for region-based image retrieval. Each image in our retrieval system is represented by a set of regions that are characterized by fuzzy sets. The FIRM measure, representing the overall similarity between two images, is defined as the similarity between two families of fuzzy sets. Compared with similarity measures based on individual regions and on all regions with crisp feature representations, our approach greatly reduces the influence of inaccurate segmentation. Experimental results based on a database of about 200,000 general-purpose images demonstrate improved accuracy, robustness, and high speed.

UR - http://www.scopus.com/inward/record.url?scp=0034792530&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0034792530&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:0034792530

SP - 543

EP - 545

Y2 - 30 September 2001 through 5 October 2001

ER -