Semantic video abstracting: Automatic generation of movie trailers based on video patterns

Abstract

Summarizing and abstracting of multimedia data is a field of great research interest, especially for multiple video platforms. In this paper we describe a system capable of generating Hollywood movie trailers automatically by using audio and video processing algorithms, combined with ontology-based knowledge and CLIPS, a rule based system. We show that the system is capable of generating convincing movie trailers for the action genre. Further work has been done to extend the results to other movie genres. © 2012 Springer-Verlag.