An Algorithmic Approach for General Video Summarization

Abstract

In the current world, multimedia has a significant role in communicating information. Videos can convey more information. Two drawbacks of the video make it inconvenient in some circumstances. First, it requires more storage. Second, we require watching a video completely to identify the content, which takes too much time. Our proposed work makes the video easy to use by solving these issues. We are trying to reduce the volume of a video by creating its summary. Summarization may either produce a image/video as output. We generate a summary video. Duplicate frame removal and stroboscopic imaging are the main techniques used in the work. For better results, the shots identified at the initial step are further processed to create their own summary clips. Summary clips are clustered together to form the final summary. The result of the proposed work is a summary video with very limited number of frames. Our proposed work can generate summary for any type of videos such as entertainment, game, surveillance and home videos. The summary video keeps the continuity of the video and conveys the meaning too.