Fusion Research For Scene Of Online3D Video Conference System

Along with the informatization of society, people’s concern on efficiencyis increasing. As hardware’s growing capacity on image processing, and thehigher speed of internet communication, online video conference system hasbecome a very important product. It has evolved from analog TV conferenceto nowadays connection specialized video conference system and IP internetvideo conference system.Given the more and more powerful PC terminal and broader bandwidth,software based video conference system will become typical application orthe case. Besides, blooming3D image technology has offered enrichedpossibility for multiple application. In this paper, we propose a Kinect camerabased multi-media online3D virtual video conference system. It’s composedof server and client terminal. Clients collect users’3D human model, send tothe server through internet connection, and the server will complete fusion forhuman model and virtual conference scene. With these work done, will see ascene where attenders gather together and have the meeting.In system construction, the paper mainly discuss several connectionmode. Through analysis on efficiency and complexity, as well as the functiondemand, it’s finalized as centralized connection. And the hard ware andsoftware configuration are completed for system realization.In the discussion for3D model fusion, a virtual conference room is builtto function as the scene, Then the cloud point data unit is optimized to lowersystem resource cost. As for the failure on traditional triangle renderingmethod, a new arrange rule, which appeals more to the triangle element onsubject surface, is implemented. As well, points rendering method and linerendering method is also proposed, due to the feature that there is such large amount of data. The result shows the3D model distracted from Kinect cameracan achieve ideal rendering effect through certain transformation in theDirect3D environment.