A Formal Model for Video Shot Segmentation and its Application via Animate Vision |
| |
Authors: | Albanese Massimiliano Chianese Angelo Moscato Vincenzo Sansone Lucio |
| |
Affiliation: | (1) Dipartimento di Informatica e Sistemistica, Università di Napoli Federico II, Italy |
| |
Abstract: | The first step in a video indexing process is the segmentation of videos into meaningful parts called shots. In this paper we present a formal model of the video shot segmentation process. Starting from a mathematical characterization of the most common transition effects, a video segmentation algorithm capable to detect both abrupt and gradual transitions is proposed. The proposed algorithm is based on the computation of an arbitrary similarity measure between consecutive frames of a video. The algorithm has been tested adopting a similarity metric based on the Animate Vision theory and results have been reported. |
| |
Keywords: | video indexing video segmentation shot transitions animate vision |
本文献已被 SpringerLink 等数据库收录! |