Key frame extraction for video content summarization using orthogonal transforms and fractional energy coefficients

Video content management system is now a day´s becoming a very popular and widely used system in multimedia retrieval. With the increase in bandwidth and internet availability lot of multimedia digital content such as movies, videos, TV shows and sports highlights are easily available. This facilitates the easy access of digital content to users. In such situation when users don´t have enough time to watch the entire video only seeks for the highlights or the important information. At this time the video summary can be helpful in fulfilling this need of users with only valuable content distribution. This is video content summarization. In this paper video content summary is prepared using key frame extraction and orthogonal transforms such as Cosine, Haar, Hartley, Walsh, Slant and Kekre transforms. The key frame extraction is done using 25%, 6.25%, 1.5625 % fractional energy coefficients of transformed video frames. Here the completeness measure is defined to check performance of each of the combination (i.e. DCT with 100 % 25% 6.25% and 1.5625%) for every transformed video clip. In the proposed method of video content summarization orthogonal transformed video gives better completeness in the form of higher completeness upto 69%.