A scalable block matching and moving Edge detection using modified kohonen Mapping based video compression

Abstract

The emerging home networks and broadband convergence network newlinetechnologies allow people to watch images and videos anywhere and anytime newlineThe display devices can be a large plasma display panel a standard definition newlineTV or a very small cellphone Consequently the video codec used in these newlinedevices should support various displays with different resolutions Hence newlinethere comes a high demand for an efficient video coding technique which can newlineprovide an acceptable quality for very low bitrate and support new newlinefunctionalities such as peak signal to noise ratio PSNR frame rate and newlinespatial scalability Moreover system should have good potential to carry newlinespatial and temporal scalability which are considered to be most essential for newlinethe new multimedia and video surveillance applications newlineIn this thesis four different fast and accurate video compression newlinetechniques are introduced to assure much reduced processing delay while newlinemaintaining good reconstruction quality Basically video signal has temporal newlineredundancies due to high correlation between successive frame Current newlinesystems fail to exploit the significant temporal redundancy in the video newlineframes to improve compression efficiency However they are ineffective in newlinemaintaining high compression ratio high gain and coding time Hence an newlineattempt has been made in the proposed algorithms to improve the newlineperformance of video compression system newline