We present several algorithms suitable for analysis of broadcast video. First, we show how wavelet analysis of frames of video can be used to detect transitions between shots in a video stream, thereby dividing the stream into segments. Next we describe how each segment can be inserted into a video database using an indexing scheme that involves a wavelet-based "signature." Finally, we show that during a subsequent broadcast of a similar or identical video clip, the segment can be found in the database by quickly searching for the relevant signature. The method is robust against noise and typical variations in the video stream, even global changes in brightness that can fool histogram-based techniques. In the paper, we compare experimentally our shot transition mechanism to a color histogram implementation, and also evaluate the effectiveness of our databasesearching scheme. Our algorithms are very efficient and run in realtime on a desktop computer. We describe how this technology could be employed to construct a "smart VCR" that was capable of alerting the viewer to the beginning of a specific program or identifying commercials and then muting the volume on the TV.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.