Live@Web.Com — Using CBIR Technology in Interactive Web-TV
The increasing amount of internet based television broadcasts has lead to new approachs to interactivity in TV programs. We developed a system which is able to supply the viewer of the program upon interaction with information relating to the program, only based on the low-level visual content of the scene. This aim is achieved by comparing signatures describing the visual content of single frames of the video with a remote database of signatures derived from known videos. The database actually links the visual information contained in the signatures to some second-level information interesting for the user. Two main problems in extending CBIR technology to videos must be overcome, one is the extraction of the visual information out of the highly redundant video material, and the other is reducing the matching time of the system enough to allow for web-based interactivity.
KeywordsVisual Content Video Retrieval Video Database Video Material Matching Time
Unable to display preview. Download preview PDF.
- 4.Macer, P, Thomas, P.: Browsing Video Content. Proceedings of BCS HCI Group Conference, The Active Web, Staffordshire University, UK., January 1999Google Scholar
- 7.Volmer, S.: Tracing Images in Large Databases by Comparison of Wavelet Fingerprints. In Proc. of the 2nd Int’l Conf. on Visual Information Systems, pp. 163–172, La Jolla, USA. December 1997.Google Scholar
- 8.Volmer, S.: Buoy Indexing of Metric Feature Spaces for Fast Approximate Image Queries. To appear in Proc. of the 6th Eurographics Workshop on Multimedia, Manchester, UK, September 2001.Google Scholar
- 10.Zhuang, Y., Rui, Y., Huang, T., Mehrotra, S.: Adaptive Key Frame Extraction Using Unsupervised Clustering. In Proc. of IEEE Int’l Conf. on Image Processing, pp. 866–870, Chicago, USA, October 1998.Google Scholar