Applying multi-class SVMs into scene image classification

Jianfeng Ren; Yuntao Shen; Songhui Ma; Lei Guo

doi:10.1007/978-3-540-24677-0_95

Applying multi-class SVMs into scene image classification

Jianfeng Ren
, Yuntao Shen
, Songhui Ma
, Lei Guo

School of Automation

Northwestern Polytechnical University Xian

Research output: Contribution to journal › Conference article › peer-review

15 Scopus citations

Abstract

Grouping images into semantically meaningful categories using the low-level visual features is a challenging and important problem in content-based image retrieval and other applications. In this paper, we show a specific high-level classification problem (scene images classification) using the low level features such as representative colors and Gabor textures. Based on the low level features, we introduce the multi-class SVMs to merge these features with the final goal to classify the different scene images. Experimental results show our method is promising.

Original language	English
Pages (from-to)	924-934
Number of pages	11
Journal	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	3029
DOIs	https://doi.org/10.1007/978-3-540-24677-0_95
State	Published - 2004
Event	17th International Conference on Industrial and Engineering Applications of Artificial Intelligence and Expert Systems, IEA/AIE 2004 - Ottowa, Ont., Canada Duration: 17 May 2004 → 20 May 2004

Access to Document

10.1007/978-3-540-24677-0_95

Cite this

@article{e6a60bb7177245d090b7aec496d7b4cc,

title = "Applying multi-class SVMs into scene image classification",

abstract = "Grouping images into semantically meaningful categories using the low-level visual features is a challenging and important problem in content-based image retrieval and other applications. In this paper, we show a specific high-level classification problem (scene images classification) using the low level features such as representative colors and Gabor textures. Based on the low level features, we introduce the multi-class SVMs to merge these features with the final goal to classify the different scene images. Experimental results show our method is promising.",

author = "Jianfeng Ren and Yuntao Shen and Songhui Ma and Lei Guo",

year = "2004",

doi = "10.1007/978-3-540-24677-0\_95",

language = "英语",

volume = "3029",

pages = "924--934",

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

issn = "0302-9743",

publisher = "Springer Science and Business Media Deutschland GmbH",

note = "17th International Conference on Industrial and Engineering Applications of Artificial Intelligence and Expert Systems, IEA/AIE 2004 ; Conference date: 17-05-2004 Through 20-05-2004",

}

TY - JOUR

T1 - Applying multi-class SVMs into scene image classification

AU - Ren, Jianfeng

AU - Shen, Yuntao

AU - Ma, Songhui

AU - Guo, Lei

PY - 2004

Y1 - 2004

N2 - Grouping images into semantically meaningful categories using the low-level visual features is a challenging and important problem in content-based image retrieval and other applications. In this paper, we show a specific high-level classification problem (scene images classification) using the low level features such as representative colors and Gabor textures. Based on the low level features, we introduce the multi-class SVMs to merge these features with the final goal to classify the different scene images. Experimental results show our method is promising.

AB - Grouping images into semantically meaningful categories using the low-level visual features is a challenging and important problem in content-based image retrieval and other applications. In this paper, we show a specific high-level classification problem (scene images classification) using the low level features such as representative colors and Gabor textures. Based on the low level features, we introduce the multi-class SVMs to merge these features with the final goal to classify the different scene images. Experimental results show our method is promising.

UR - https://www.scopus.com/pages/publications/9644290929

U2 - 10.1007/978-3-540-24677-0_95

DO - 10.1007/978-3-540-24677-0_95

M3 - 会议文章

AN - SCOPUS:9644290929

SN - 0302-9743

VL - 3029

SP - 924

EP - 934

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

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

T2 - 17th International Conference on Industrial and Engineering Applications of Artificial Intelligence and Expert Systems, IEA/AIE 2004

Y2 - 17 May 2004 through 20 May 2004

ER -

Applying multi-class SVMs into scene image classification

Abstract

Access to Document

Other files and links

Fingerprint

Cite this