Video Information Retrieval (2)
- Zhang, et al, "Video Parsing, Retrieval and Browsing: An Integrated and Content-Based Solution," 1996
- Video is segmented into individual shots, (95% precision)
- Extract 1 or 2 key frames from each shot (1 frame for only one scene in a shot; 2, the first and the last frame for indicating change of the scene in a shot)
- Color, texture, shape are extracted from key frames to allow QBIC
- Provide hierarchical browsing interface for reducing transmission bandwidth
- Example:
http://www.iss.nus.sg/RND/MS/Projects/vc/vidorigin.html
Notes: