Video abstract extraction using semantic categorization

Semantic techniques can be utilized for understanding the meaning of sentences in a text. We can refer the abstract of some document in order to understand its contents very quickly. But the case of video data is different. We cannot review the whole video file in a short time period just like the text document case. This paper proposes the design of video abstract extraction. In order to generate the abstract it also defines semantic categorization of the video. Once the category of the video is accurately determined then only those sentences can be selected that relate with the category of the video in order to prepare the video abstract.