| With the wide range of applications and concerns of the stereo video technology,multi-view video technology has also become the focus of the video media technology.There is a huge difference between the multi-view video media and ordinary video mediaon the amount of data, which restricts the development and wide application of themulti-view video. Obviously, multi-view video coding and transmission is currently thefocus of the study, so this article researches these two directions.Efficient video coding is an important guarantee of the video transmission.Withoutfast video compression, there is no way to guarantee the real-time transmission ofmulti-view video on the network. Due to the multi-viewpoint video characteristics ofcarrying large amounts of data information, its encoding time is too long. This paper, theproblem of the transmission-oriented multi-view video reduces encoding time. In the caseof the expense of minimum PSNR search improvements of the proposed video coding is,significantly reduce the compression time, and achieve the purpose of ease of networktransmission. For the process of transmission of the video the error control is the assuranceof the transmission.According to achieve due to the large amount of data of the multi-view video, thevideo transmission adds redundant error compared to the control information, reduce thetime of encoding prior to transmission error concealment technique, adds real-time networktraffic, reduces the amount of data transmitted, and reduces the pressure of the networkbandwidth compared to the control information in coding side in the error control of thevideo transmission.Research on the multi-view video, we propose on inter-frame motion vector of themotion vector comparison between multi-viewpoint video viewpoint single viewpoint ofthe three properties:1. An image the number of macro blocks having a motion vector of the share of thetotal number of macro blocks the increase in the proportion.2. The adjacent viewpoint motion vector is relatively simple, the majority of thehorizontal and vertical directions. 3. The absolute value of the motion vectors are relatively small.These three characteristics play a significant role.In this paper, the vector prediction, the termination conditions and the search templatein the EPZS search algorithm has been improved. In the improved process, the paper isbased on reducing the encoding time for facilitate network transmission purpose. In thevector prediction, a high proportion of the prediction is the horizontal and verticaldirections. Termination condition simplifies the termination condition computationalcomplexity. Provide a more focused search template to search in the search template.Search algorithm optimized by experiment in a small decrease in PSNR value of the imagereduce a large number of the search time, achieve the purpose of reducing encoding timeand facilitate real-time transmission of the network.In this paper, the error concealment of the decoding is used to the multi-viewpointvideo transmission error control.Contrasting to the error control scheme of the error controlinformation added to the encoding side, this program effectively remove redundantinformation, reduce the amount of data transmitted, improve the transmission efficiency,reduce encoding time, and achieve good control effect. In this paper, the process of errorconcealment is completed by scan frame, determining the error macro-block, selectingreference frame, determining the mask sequence, obtaining the motion vector. The PSNRvalue of the video image obtained by experiment can be improved. The decoding end errorconcealment scheme has better image recovery capability.The use of the two methods described above can effectively reduce encoding time,improve the coding efficiency, reduce redundant information, reduce the amount of datatransmitted, improve transmission efficiency, and optimize the storage and transmission ofmulti-viewpoint video. |
PDF Full Text Request