The mushroom growth of video information, consequently, necessitates the progress of content-based video analysis techniques. Video summarization, aiming to provide a short video summary of original video document, has drawn more and more attention these years. In this paper, we propose an algorithm for video summarization based on feature comparison. Video redundant information is removed and essential content is stored based on content analysis. The Impact Factor (IF ) of scene, shot and key frame is defined for video summary construction. Results of experiment applications are given to illustrate the proposed algorithm.