Title :
SVM Based Classification of Seven Nature Objects for Anytime, Anywhere Digital Photo Annotation
Author :
Song, Chull Hwan ; Yoo, Seong Joon
Author_Institution :
Sejong Univ., Seoul
Abstract :
This paper proposes a method that can be utilized for automatically annotating digital photos anytime, anywhere. A digital camera or an annotation server connected to the digital camera through a ubiquitous computing network can automatically annotate captured photos using the proposed method. Annotating digital images is not a new research problem. We have developed a novel method of classifying seven nature objects from digital images. Thus, this paper describes the method and shows that it is superior to previous methods of classifying nature objects.
Keywords :
image classification; support vector machines; ubiquitous computing; video cameras; SVM based classification; automatic digital photo annotation; digital camera; digital image annotation; nature object classification; ubiquitous computing network; Digital cameras; Digital images; Histograms; Humans; Internet; Layout; Principal component analysis; Support vector machine classification; Support vector machines; Ubiquitous computing;
Conference_Titel :
Multimedia and Ubiquitous Engineering, 2007. MUE '07. International Conference on
Conference_Location :
Seoul
Print_ISBN :
0-7695-2777-9
DOI :
10.1109/MUE.2007.199