Automatic classification of TV sports news video by multiple subspace method
✍ Scribed by Yoshiaki Sugiyama; Yasuo Ariki
- Book ID
- 101298088
- Publisher
- John Wiley and Sons
- Year
- 2000
- Tongue
- English
- Weight
- 137 KB
- Volume
- 31
- Category
- Article
- ISSN
- 0882-1666
No coin nor oath required. For personal study only.
✦ Synopsis
In video images such as those in sports news, there are constraints on the position of the objects in the image and other aspects. In other words, there are typical images that characterize individual sports, and the stories in the sports news can be classified into sports categories such as tennis and baseball. This paper proposes a method of sports news story classification based on the comprehensive features of the scene framework, without clearly identifying the physical objects in the image. The multiple subspace method is used in the classification, and a story classification rate of 98.6% is realized.