| Summary: | 碩士 === 國立臺灣大學 === 資訊工程學研究所 === 92 === With the advance of digital video analysis and storage technologies, also in the filed of entertainment industry, movie viewers hope to gain more control over what they see. For example, we expect interactivity in films produced for DVD and online entertainment-on-demand systems. Therefore, tools that enable automated movie analysis are becoming indispensable to efficiently access, retrieve, and browse information closed to a human perceptive and semantic level. We proposed an action movie segmentation and summarization system based on tempo, which is regarded as the delivery speed of important segments of a movie. In this tempo-based system, we combine the techniques of shot change detection, motion activity analysis, semantic context detection based on audio features, and the film domain knowledge (film grammar) to perform the concept of tempo and construct the system for action movies segmentation and summarization. Then we do experiments on several different action movie sequences, and give an analysis and comparison according to the experimental results.
|
|---|
