Extracting semantic information through illumination classification

Author

Aner-Wolf, Aya

Author_Institution

Dept. of Comput. Sci. & Appl. Math, Weizmann Inst. of Sci., Rehovot, Israel

Volume

1

fYear

2004

fDate

27 June-2 July 2004

Abstract

A system is presented for automatic classification of the atmosphere in motion pictures scenes. This classification is based solely on the extraction of approximate illumination information from single frames. By using well defined film grammar rules regarding scene lighting, it generates high level semantic descriptors for frames and scenes. These illumination descriptors could be further combined with other descriptors like motion and sound, to generate a complete automatic semantic analysis system of video media.

Keywords

cinematography; feature extraction; grammars; image classification; image motion analysis; image retrieval; video signal processing; vocabulary; approximate illumination information extraction; atmosphere; automatic classification; automatic semantic analysis system; cinematography; film grammar rules; high level semantic descriptors; illumination classification; illumination descriptors; motion pictures scenes; scene lighting; semantic information extraction; video media; Atmosphere; Brightness; Computer Society; Computer vision; Data mining; Distributed computing; Layout; Lighting; Pattern recognition; Reflectivity;

fLanguage

English

Publisher

ieee

Conference_Titel

Computer Vision and Pattern Recognition, 2004. CVPR 2004. Proceedings of the 2004 IEEE Computer Society Conference on

ISSN

1063-6919

Print_ISBN

0-7695-2158-4

Type

conf

DOI

10.1109/CVPR.2004.1315042

Filename

1315042