Research On3d Video Transmission Based On IPTV

IPTV take the phone, cable TV and broadband realize for a three-net fusion, it can offer avariety of digital interactive network services for home users. At the same time, the interest of3D entertainment applications from providers and users is more and more strong. Therefore,we need a new technology that transfer the3D content to the users’ terminal.In the traditional3D video transmission platform, it is mostly transmission a eyes-video,all views of the eyes video are coded singlely, and at the same time, use the traditionalclient/server way to transmission. In this platform, limiting the customers on the views offreedom of choice and network interactive, and, using the traditional client/server mode3Dvideo, the consumption of bandwidth is large. In the3D video transmission structure of thispaper, there are four modules: collecting/coding part, transmission parts, the client, and thedisplay parts. In the collecting/coding part, for every visual angle, right and left eye viewangle from the perspective of two way collection, at the same time, using layed codingmethod. In the transmission part, different from the client/server model, use Peer-to-Peermode for transmission. In the client part, using the prediction of the users’ future headposition, and also support for two display forms.In this paper, in order to better use IPTV interaction characteristics, proposes a newclient-driven from the perspectives of multiview video streaming system, it allows customerswatch3D video interactively, according to the customers’ head position, through sending asmall number of views to reduce bandwidth requirements. The user’s head position be trackedand predicted into the future to choice the view that best match the user’s current view angle.Need to predict the future position of the head, so that views matching prediction can beprefetched to avoid the delay produced by network transport and stream switching. Thissystem allocates more bandwidth to the choosed views, in order to provide the current viewangle. Highly compressed, low quality of some of the other angle of view version will also beprefetched for concealment, if the current user’s point of view different from the prediction.An objective method based on the head movement and the delay of the system is introducedto determine the number of additional lower quality views to be prefetched. The system useslayered coding method, based on the concept of multiview coding (MVC) and scalable videocoding (SVC) to improve the efficiency of compression and flexible to choose from the point of view of the bandwidth allocation.At the same time, the characteristic of intensive data of multiview video prevent itwidely spread on the network, because in traditional client/server architecture, streamingmedia server bandwidth consumption is large. As P2P streaming media technology rapiddevelopment, in the Internet over P2P networks provide multiview video steaming servicesappeared, when transmitting a lot of multiview video data, it can effectively use the users’uploading bandwidth resources to reduce the streaming media server’s bandwidthconsumption as a scalable solution on the Internet from video services provide. However, theview switching of multiview video introduce great dynamic to the system, this kind ofdynamic causes large view switching delay and low quality of some views. In order to solvethese problems, we use a half-stable structure in P2P streaming media, and organize theviewers of different views together, cooperate in views switching and content transmission, toreduce the view switching delay, and balance the bandwidth resources between views. In thispaper, through improving the half-stable structure in multiview P2P, we establishe ahalf-stable structure in the3D video transmission, when switch view point in multiviewstereoscopic video, reducing the switching delay by this structure.During the experiment, prove the proposed system is better in distortion rate, comparedwith the system that all views are coded singlely. At the same time, use a half-stable structureof3D video, when switching view point, it can effectively reduce the switching delay. Thispaper only manage the transmission of video content, without taking into account thetransmission process of audio content, and in the the process of using P2P transmission, notconsider the block scheduling problem. Therefore, the solution of the two problem is futureresearching. Believe that after the research and continue learning, researching on the3Dvideo based on IPTV transmission will be more perfect.