Key frame extraction from scenery video
Min Zhi · 2007
Key frames are representative frames in a shot. With key frames, video information can be efficiently managed in retrieval and indexing. In view of the scenery video characteristic and the user attention focus, this paper proposed a new semantic based key frame extraction method. First according to the amplitude and the angle of P frame’s motion vector, the shot motion is judged; and then according to the speed of camera motion, the duration of a shot and the distribution of motion vector, the key frame is extracted. The experiment result shows that key frame can reflect the video content well.