Key Frame Generation to Generate Activity Strip Based on Similarity Calculation

Abstract

Management of video data is done for several purposes, such as to make the information more meaningful. Research has been conducted to manage the video in terms of detecting activity in a video. There are three stages to generate activity strip: the data source stage (preparation of the frames), the processing stage (analysis of the activity), and the final stage (the collection of key frames). The generation of activity strip is done by calculating the difference of the pixel values of two frames to detect a similarity. In this research, we used SAD (Sum of Absolute Difference) method to calculate the value of the difference of the frame. Similar frames can be grouped in the same cluster. Each cluster is considered as one frame (or multiple frames) to serve as a key frame. The key frames are used for the representation of the activity strip. A collection of activity strip will be arranged sequentially and continuously for the activity generation.