Details

Autor(en) / Beteiligte

• Ren, Jianfeng
• Shen, Yuntao
• Ma, Songhui
• Guo, Lei

Titel

Applying Multi-class SVMs into Scene Image Classification

Ist Teil von

• Innovations in Applied Artificial Intelligence, 2004, p.924-934

Ort / Verlag

Berlin, Heidelberg: Springer Berlin Heidelberg

Erscheinungsjahr

2004

Link zum Volltext

Quelle

Alma/SFX Local Collection

Beschreibungen/Notizen

• 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.