Extracting semantic information in geographic image data: Some preliminary results

Rong Zhao; William I. Grosky

Extracting semantic information in geographic image data: Some preliminary results

Rong Zhao
, William I. Grosky

Center for Excellence in Wireless Technology

University of Michigan, Dearborn

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

Abstract

In this paper we present the preliminary results of our study that seeks to find semantic information in geographic image data. We applied several visual feature indexing techniques including texture, structured color texture, color anglogram, and shape anglogram to represent image contents. Our study also involves a dimension reduction technique called latent semantic indexing which has been used in textual information retrieval for years. Experimental results show that these techniques are able to uncover the underlying semantic structure of image contents. The objective of this research is to develop an integrated retrieval, analysis, and annotation framework with these techniques and to generate more meaningful metadata models for geographic image data.

Original language	English
Pages	59-65
Number of pages	7
State	Published - 2003
Event	Proceedings of the International Conference on Imaging Science, Systems and Technology, CISST'03 - Las Vegas, NV, United States Duration: Jun 23 2003 → Jun 26 2003

Conference

Conference	Proceedings of the International Conference on Imaging Science, Systems and Technology, CISST'03
Country/Territory	United States
City	Las Vegas, NV
Period	06/23/03 → 06/26/03

Keywords

Anglogram
Color
Content-based image retrieval
Geographic image
Latent semantic indexing
Texture

Cite this

@conference{1afa7955deca42f3975be81a7fc7e639,

title = "Extracting semantic information in geographic image data: Some preliminary results",

abstract = "In this paper we present the preliminary results of our study that seeks to find semantic information in geographic image data. We applied several visual feature indexing techniques including texture, structured color texture, color anglogram, and shape anglogram to represent image contents. Our study also involves a dimension reduction technique called latent semantic indexing which has been used in textual information retrieval for years. Experimental results show that these techniques are able to uncover the underlying semantic structure of image contents. The objective of this research is to develop an integrated retrieval, analysis, and annotation framework with these techniques and to generate more meaningful metadata models for geographic image data.",

keywords = "Anglogram, Color, Content-based image retrieval, Geographic image, Latent semantic indexing, Texture",

author = "Rong Zhao and Grosky, \{William I.\}",

year = "2003",

language = "English",

pages = "59--65",

note = "Proceedings of the International Conference on Imaging Science, Systems and Technology, CISST'03 ; Conference date: 23-06-2003 Through 26-06-2003",

}

TY - CONF

T1 - Extracting semantic information in geographic image data

T2 - Proceedings of the International Conference on Imaging Science, Systems and Technology, CISST'03

AU - Zhao, Rong

AU - Grosky, William I.

PY - 2003

Y1 - 2003

N2 - In this paper we present the preliminary results of our study that seeks to find semantic information in geographic image data. We applied several visual feature indexing techniques including texture, structured color texture, color anglogram, and shape anglogram to represent image contents. Our study also involves a dimension reduction technique called latent semantic indexing which has been used in textual information retrieval for years. Experimental results show that these techniques are able to uncover the underlying semantic structure of image contents. The objective of this research is to develop an integrated retrieval, analysis, and annotation framework with these techniques and to generate more meaningful metadata models for geographic image data.

AB - In this paper we present the preliminary results of our study that seeks to find semantic information in geographic image data. We applied several visual feature indexing techniques including texture, structured color texture, color anglogram, and shape anglogram to represent image contents. Our study also involves a dimension reduction technique called latent semantic indexing which has been used in textual information retrieval for years. Experimental results show that these techniques are able to uncover the underlying semantic structure of image contents. The objective of this research is to develop an integrated retrieval, analysis, and annotation framework with these techniques and to generate more meaningful metadata models for geographic image data.

KW - Anglogram

KW - Color

KW - Content-based image retrieval

KW - Geographic image

KW - Latent semantic indexing

KW - Texture

UR - https://www.scopus.com/pages/publications/1842532949

M3 - Paper

AN - SCOPUS:1842532949

SP - 59

EP - 65

Y2 - 23 June 2003 through 26 June 2003

ER -

Extracting semantic information in geographic image data: Some preliminary results

Abstract

Conference

Keywords

Other files and links

Fingerprint

Cite this