Semi-automatic Video Content Annotation

Video modeling and annotating are indispensable operations necessary for creating and populating a video database. To annotate video data effectively and accurately, a video content description ontology is first proposed in this paper, we then introduce a semi-automatic annotation strategy which utilize various video processing techniques to help the annotator explore video context or scenarios for annotation. Moreover, a video scene detection algorithm which joints visual and semantics is proposed to visualize and refine the annotation results. With the proposed strategy, a more reliable and efficient video content description could be achieved. It is better than manual manner in terms of efficiency, and better than automatic scheme in terms of accuracy.