Design And Realization Of Custom Data Reduction Video Playback System Based On Red5

As computer performance continues to improve and the continuous developmentof stream media technology，people put forward higher requirements on the Internetvideo playback，such as in the area of education, provides users with more than onecourse preview. A Terminal displaying multiple images at the same time, eachshowed a video content, this requires synthesis to multi-screen video processing. Pastmultiple images with the main application in video conference or in the monitoringand control system, but the video is in a fixed number and size of each image. And inthe case of limited bandwidth, some users demand low video quality, no need for itsentire output, waste of bandwidth. How the user can according to their own needs toselect and configure the video content,appropriate simplify the video data,and makefull use of bandwidth transfer multi-screen video for playbacking, for these issues,current video playback system did not have any relevant research and matureapplication.In this context, firstly analysis and research the current video-image synthesismethod, proposes a multi-image synthesis method with the pixels of image based onframes restructuring, support for multiple video formats, such as in the field ofstreaming media of the most widely used FLV format, and the user can according totheir own needs to customize the screen content, simplify video data, achieve thepurpose of rational use of bandwidth. Both Red5streaming server as a core system fornetwork video player, integrated video-image synthesis, expand its video processingcapabilities. While designing and implementing a client control to the server-sidefeedback of users custom requirements, reduce resource consumption, rapid bufferingand playback on a limited bandwidth.The current mainstream blog platformWordPress combined as application scenarios, provides a new video servicesolution for users, and also for the blog to provide a better user experience for richmedia applications(RIA). Finally through experimental verificate the feasibility of themulti-image synthesis methods presented in this article，and analysis the applicationeffect on the whole platform.