Extracting Story Units from Video Using Contextual Information

In this paper, we propose a story unit extraction method using contextual information in video shots. We divide the contextual information into two categories: local and global contextual information. We represent the local contextual information in the video shot as foreground region's information, shot activity, and background color semantics. The global contextual information is determined by local contextual similarity and time locality. It also takes into account the effect of surrounding or neighboring shots. Using contextual information, we extract desirable story unit boundaries from video shots.

• Type: Conference
• Year: 2000
• Where: MVA
• Authors: Hang-Bong Kang
• Keywords: Contextual Information, Global Contextual Information, MVA 2000, MVA 2007, Video Shots