Video Summarization by Motion Analysis: Using Optical Flow Technique
Mini Zeng, Guang Qiu Huang · 2011
Video summaries aim to represent all the frames in a video. They can rapidly provide users with information about the content of a large video or set of videos. And key frames are often arranged as storyboards, in which the key frames represent shots and sequences to summarize the story. Storyboards are often used by directors to plan films. More important, if key frames are properly selected, many image content searches can be conducted on the key frames rather than the complete video, greatly reducing the computational requirements for search. In this issue, We get key frames from a selected video by using Lucas Kanade Optical Flow Technique to compute motion of optical flow of key elements in each frames. then analyze the curve graph by Max Motion Algorithm and Extra Maximum Motion Algorithm to get key frames.