Evaluation of color representation for texture analysis

R. Verbrugge; Egon van den Broek; E.M. van Rikxoort; N. Taatgen; L. Schomaker

Evaluation of color representation for texture analysis

R. Verbrugge (Editor), Egon van den Broek, E.M. van Rikxoort, N. Taatgen (Editor), L. Schomaker (Editor)

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

183 Downloads (Pure)

Abstract

Since more than 50 years texture in image material is a topic of research. Hereby, color was ignored mostly. This study compares 70 different configurations for texture analysis, using four features. For the configurations we used: (i) a gray value texture descriptor: the co-occurrence matrix and a color texture descriptor: the color correlogram, (ii) six color spaces, and (iii) several quantization schemes. A three classifier combination was used to classify the output of the configurations on the VisTex texture database. The results indicate that the use of a coarse HSV color space quantization can substantially improve texture recognition compared to various other gray and color quantization schemes.

Original language	Undefined
Pages	35-42
Number of pages	8
Publication status	Published - 21 Oct 2004
Event	16th Belgium-Dutch Conference on Artificial Intelligence, BNAIC 2004 - Groningen, Netherlands Duration: 21 Oct 2004 → 22 Oct 2004 Conference number: 16

Conference

Conference	16th Belgium-Dutch Conference on Artificial Intelligence, BNAIC 2004
Abbreviated title	BNAIC
Country/Territory	Netherlands
City	Groningen
Period	21/10/04 → 22/10/04

Keywords

HMI-VRG: Virtual Reality and Graphics
Texture
Pattern Recognition
co-occurence matrix
color correlogram
Color
HMI-CI: Computational Intelligence
HMI-HF: Human Factors
IR-79317
Classification
Evaluation
EWI-21117

Access to Document

http://irs.ub.rug.nl/dbi/4357a847c1590

Cite this

@conference{b3cab655ccfa45d3af08cf5994649789,

title = "Evaluation of color representation for texture analysis",

abstract = "Since more than 50 years texture in image material is a topic of research. Hereby, color was ignored mostly. This study compares 70 different configurations for texture analysis, using four features. For the configurations we used: (i) a gray value texture descriptor: the co-occurrence matrix and a color texture descriptor: the color correlogram, (ii) six color spaces, and (iii) several quantization schemes. A three classifier combination was used to classify the output of the configurations on the VisTex texture database. The results indicate that the use of a coarse HSV color space quantization can substantially improve texture recognition compared to various other gray and color quantization schemes.",

keywords = "HMI-VRG: Virtual Reality and Graphics, Texture, Pattern Recognition, co-occurence matrix, color correlogram, Color, HMI-CI: Computational Intelligence, HMI-HF: Human Factors, IR-79317, Classification, Evaluation, EWI-21117",

author = "R. Verbrugge and {van den Broek}, Egon and {van Rikxoort}, E.M. and N. Taatgen and L. Schomaker",

year = "2004",

month = oct,

day = "21",

language = "Undefined",

pages = "35--42",

note = "null ; Conference date: 21-10-2004 Through 22-10-2004",

}

Evaluation of color representation for texture analysis. / Verbrugge, R. (Editor); van den Broek, Egon; van Rikxoort, E.M.; Taatgen, N. (Editor); Schomaker, L. (Editor).

2004. 35-42 Paper presented at 16th Belgium-Dutch Conference on Artificial Intelligence, BNAIC 2004, Groningen, Netherlands.

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

TY - CONF

T1 - Evaluation of color representation for texture analysis

AU - van den Broek, Egon

AU - van Rikxoort, E.M.

A2 - Verbrugge, R.

A2 - Taatgen, N.

A2 - Schomaker, L.

N1 - Conference code: 16

PY - 2004/10/21

Y1 - 2004/10/21

N2 - Since more than 50 years texture in image material is a topic of research. Hereby, color was ignored mostly. This study compares 70 different configurations for texture analysis, using four features. For the configurations we used: (i) a gray value texture descriptor: the co-occurrence matrix and a color texture descriptor: the color correlogram, (ii) six color spaces, and (iii) several quantization schemes. A three classifier combination was used to classify the output of the configurations on the VisTex texture database. The results indicate that the use of a coarse HSV color space quantization can substantially improve texture recognition compared to various other gray and color quantization schemes.

AB - Since more than 50 years texture in image material is a topic of research. Hereby, color was ignored mostly. This study compares 70 different configurations for texture analysis, using four features. For the configurations we used: (i) a gray value texture descriptor: the co-occurrence matrix and a color texture descriptor: the color correlogram, (ii) six color spaces, and (iii) several quantization schemes. A three classifier combination was used to classify the output of the configurations on the VisTex texture database. The results indicate that the use of a coarse HSV color space quantization can substantially improve texture recognition compared to various other gray and color quantization schemes.

KW - HMI-VRG: Virtual Reality and Graphics

KW - Texture

KW - Pattern Recognition

KW - co-occurence matrix

KW - color correlogram

KW - Color

KW - HMI-CI: Computational Intelligence

KW - HMI-HF: Human Factors

KW - IR-79317

KW - Classification

KW - Evaluation

KW - EWI-21117

M3 - Paper

SP - 35

EP - 42

Y2 - 21 October 2004 through 22 October 2004

ER -