Morphological Color Quantization

S. Gibson; R. W. Harvey

doi:10.1109/CVPR.2001.991007

Morphological Color Quantization

S. Gibson, R. W. Harvey

Research output: Contribution to conference › Paper

3 Citations (Scopus)

Abstract

Color histograms are a central feature in many image retrieval systems. Indeed they are part of the MPEG-7 standard. But histograms suffer from the "curse of dimensionality" in which the number of bins increases exponentially with the number of dimensions. There is therefore an imperative for methods for simplifying histograms. This paper presents a new method for simplifying histograms based on a cascade of increasing-scale graph morphology filters. The system we choose preserves scale space causality and so preserves the modes of the histogram. The method is quick to compute so is therefore a practically useful feature. We present results using the MPEG-7 Common Color Dataset that show that these new compressed features have a retrieval performance that is equivalent to full histograms.

Original language	English
Pages	525-530
Number of pages	6
DOIs	https://doi.org/10.1109/CVPR.2001.991007
Publication status	Published - Dec 2001
Event	2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition - Kauai, United States Duration: 8 Dec 2001 → 14 Dec 2001

Conference

Conference	2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition
Country/Territory	United States
City	Kauai
Period	8/12/01 → 14/12/01

Access to Document

10.1109/CVPR.2001.991007

Cite this

@conference{dd01b5a2c23c422bb0f7b71e141e0d1c,

title = "Morphological Color Quantization",

abstract = "Color histograms are a central feature in many image retrieval systems. Indeed they are part of the MPEG-7 standard. But histograms suffer from the {"}curse of dimensionality{"} in which the number of bins increases exponentially with the number of dimensions. There is therefore an imperative for methods for simplifying histograms. This paper presents a new method for simplifying histograms based on a cascade of increasing-scale graph morphology filters. The system we choose preserves scale space causality and so preserves the modes of the histogram. The method is quick to compute so is therefore a practically useful feature. We present results using the MPEG-7 Common Color Dataset that show that these new compressed features have a retrieval performance that is equivalent to full histograms.",

author = "S. Gibson and Harvey, {R. W.}",

year = "2001",

month = dec,

doi = "10.1109/CVPR.2001.991007",

language = "English",

pages = "525--530",

note = "2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition ; Conference date: 08-12-2001 Through 14-12-2001",

}

TY - CONF

T1 - Morphological Color Quantization

AU - Gibson, S.

AU - Harvey, R. W.

PY - 2001/12

Y1 - 2001/12

N2 - Color histograms are a central feature in many image retrieval systems. Indeed they are part of the MPEG-7 standard. But histograms suffer from the "curse of dimensionality" in which the number of bins increases exponentially with the number of dimensions. There is therefore an imperative for methods for simplifying histograms. This paper presents a new method for simplifying histograms based on a cascade of increasing-scale graph morphology filters. The system we choose preserves scale space causality and so preserves the modes of the histogram. The method is quick to compute so is therefore a practically useful feature. We present results using the MPEG-7 Common Color Dataset that show that these new compressed features have a retrieval performance that is equivalent to full histograms.

AB - Color histograms are a central feature in many image retrieval systems. Indeed they are part of the MPEG-7 standard. But histograms suffer from the "curse of dimensionality" in which the number of bins increases exponentially with the number of dimensions. There is therefore an imperative for methods for simplifying histograms. This paper presents a new method for simplifying histograms based on a cascade of increasing-scale graph morphology filters. The system we choose preserves scale space causality and so preserves the modes of the histogram. The method is quick to compute so is therefore a practically useful feature. We present results using the MPEG-7 Common Color Dataset that show that these new compressed features have a retrieval performance that is equivalent to full histograms.

U2 - 10.1109/CVPR.2001.991007

DO - 10.1109/CVPR.2001.991007

M3 - Paper

SP - 525

EP - 530

T2 - 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition

Y2 - 8 December 2001 through 14 December 2001

ER -