Video Annotation Based on Video Structure Information and Key Frames Semantic Description

Video annotation is a kind of "high-level feature extraction" or "semantic concept detection", which is a promising approach to bridging semantic gap between low-level features and user descriptions. In this paper we propose a two layer video annotation scheme based on video structure and the visual context information through the video clips. To improve the retrieval performance based on annotation text we combine the video title and related text information to re-rank the retrieval sequences. In the experiment we test our annotation result through text based video retrieval experiment meanwhile we presented our re-rank result. Our annotation and re-ranking approach could be utilized in future web video data organization.