V.Joseph Emmanuvel, S. Vetrivel
Existing Video Indexing Models are analyzed and a practical approach to the optimal Video Indexing is introduced. It is studied under all the phases of video indexing processes like segmentation, indexing, database storage, query based access, browsing and video clip retrieval, etc… The main aim is to easily parse the video stream into meaningful scenes, maintain them in an effective database with minimal data repetitions, efficient query handling and user friendly browsing capabilities. Conceptual Graph, Motion Estimation, Mean Absolute Frame Difference, Displaced Frame Difference, Dublin Core and other important existing techniques are utilized in this model. The main aim is to reduce the memory storage of video clippings without visible loss in quality by using a predictive video compression technique. Today almost all video clippings face a compromise between their quality and memory size. Even Video clippings are not advised to include in the web pages because of their downloading time and its memory size. To try to rectify it and to take positive measures to convert a video clipping file similar to *.swf (flashplayer’s schockwave files) is the aim of this presentation.
Video Compression, Content Based Video