부산대학교 도서관

News videos are a valuable tool for disseminating information. However, the number of newscasts every day on a single channel is enormous, and when multiplied by the number of news channels, the volume of information transmitted each day is enormous. In such instances, sorting news videos is critical for creating a well-organized database of news videos. Topics like Video Segmentation, Video Indexing, and Retrieval of the specific news from the news videos are the trending topics and points of interest for researchers from all over the world as these topics have various applications. This paper discusses segmentation of videos, as well as indexing and retrieval strategies of the videos, focusing on some of the methods for structure analysis of videos, such as SBD (Shot Boundary Detection), scene segmentation, story segmentation, key frame extraction, commercial filtering, and speech recognition. Finally, we look at potential research directions in the future.