1. Introduction


1. Introduction

Automatic video summarization requires characterization and a means to rank or select a small subset of a video as summary output. For accurate characterization, techniques must be used that incorporate audio, image and language features from video. Multimodal analysis uses different features to achieve improved results over a single mode of data. A single modal feature description may be combined or used individually to select video for summarization.

The accuracy of a video summary is often based on subjective matters such as ease-of-use, visual display and user control. User-studies and feedback surveys are usually the preferred means to assess an accuracy or quality measure to the final result. Visualization and display are also important elements in creating video summaries. A summary system may accurately locate the most important areas of a video, but the display of the results must be shown in a manner that is useful to the viewer.

The remainder of the chapter is presented as follows: Section 2, A description of image, audio, language and other video features; Section 3, Video surrogates and the process of combining video features for characterization; Section 4, Methods for evaluation and user-studies; and Section 5, Visualization techniques for summaries.




Handbook of Video Databases. Design and Applications
Handbook of Video Databases: Design and Applications (Internet and Communications)
ISBN: 084937006X
EAN: 2147483647
Year: 2003
Pages: 393

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net