Automatic Categorization of Shots in News Videos Based on the Temporal Relations

摘要

The development of new methods and technologies of video indexing and retrieval is stimulated by the growing amount of digital video data stored in Internet video collections, TV shows archives, video-on-demand systems, personal video archives offered by Web services, etc. The videos very frequently offered in the Web by broadcast channels are news videos and sports news videos. Content-based indexing of videos is based on the automatic detection of a video structure. A video shot is the main structural video unit. Shots can be of different categories such as intro or final animation, chart or table shots, anchor, reporter, statement, or interview shots, and finally the most informative report shots. The temporal aggregation results in grouping of shots into scenes of a given category. The paper examines the usefulness of the temporal aggregation method to select report shots and non-report shots in a news video.