Video Summarization for Surveillance System Using key-frame Extraction based on Cluster

Abstract

The amount of data has grown in recent years due to the use of a vast number of videos, which requires time to access them in addition to the difficulty of browsing and retrieving the video content. To fix this issue, it was proposed that the videos be summarized for easy access and that the content of the videos is browsed easier. The primary objective of the video summary is to provide a simple description of the video by removing the redundancy and extracting keyframes from the video. This paper will clarify the four ways that are using to summing up the video based on the keyframe extraction. In frames extraction, the first two methods rely on the threshold value, while the second two methods rely on clustering to extract the keyframes.