本文選題：H.265 + MPEG-DASH��； 參考：《武漢理工大學(xué)》2014年碩士論文

【摘要】：隨著網(wǎng)絡(luò)不斷的建設(shè)以及超清、高清視頻數(shù)量的增加，流媒體業(yè)務(wù)在整個(gè)網(wǎng)絡(luò)流量中占的比重不斷提高，由此也使國(guó)內(nèi)外對(duì)流媒體技術(shù)進(jìn)行不斷的研究，針對(duì)目前流媒體解決方案的缺點(diǎn)，本文提出了一個(gè)由H.265和MPEG-DASH相結(jié)合的流媒體解決方案。 本文首先分析了高效視頻壓縮標(biāo)準(zhǔn)H.265的關(guān)鍵技術(shù)，探討了H.265視頻質(zhì)量評(píng)價(jià)相關(guān)內(nèi)容，選取PSNR和SSIM兩個(gè)參數(shù)對(duì)H.265視頻質(zhì)量進(jìn)行客觀(guān)評(píng)價(jià)。其次研究了MPEG-DASH相關(guān)技術(shù)，對(duì)目前的自適應(yīng)算法進(jìn)行了分析，結(jié)合H.265視頻質(zhì)量評(píng)價(jià)相關(guān)內(nèi)容提出了基于視頻質(zhì)量的自適應(yīng)算法，算法在基于帶寬的自適應(yīng)算法基礎(chǔ)上在碼流切換過(guò)程中考慮到視頻質(zhì)量和接收緩沖的使用量及其變化量問(wèn)題，利用視頻質(zhì)量可以改善碼流切換過(guò)程畫(huà)面變化過(guò)大的問(wèn)題。對(duì)于緩沖區(qū)的考慮，在綜合以前算法的基礎(chǔ)上主要考慮在緩沖使用過(guò)高或過(guò)低時(shí)碼流的切換問(wèn)題，用matlab平臺(tái)實(shí)現(xiàn)了基于視頻質(zhì)量的自適應(yīng)算法，對(duì)算法的功能進(jìn)行了仿真驗(yàn)證，改進(jìn)算法在緩沖區(qū)內(nèi)媒體較少、過(guò)多、網(wǎng)絡(luò)帶寬變化較大的情況下性能優(yōu)于基于帶寬的自適應(yīng)算法。 本文在最后結(jié)合了H.265和MPEG-DASH實(shí)現(xiàn)了一個(gè)基于libdash流媒體系統(tǒng)，系統(tǒng)使用MPEG-DASH作為傳輸協(xié)議，服務(wù)器端使用H.265來(lái)對(duì)視頻進(jìn)行壓縮，客戶(hù)端以libdash庫(kù)為基礎(chǔ)采用多線(xiàn)程技術(shù)，實(shí)現(xiàn)了基于帶寬的自適應(yīng)算法，在局域網(wǎng)和廣域網(wǎng)環(huán)境中對(duì)6種場(chǎng)景下的系統(tǒng)進(jìn)行測(cè)試，測(cè)試結(jié)果可以看出系統(tǒng)能夠?qū)τ蒆.265壓縮的dash流進(jìn)行傳輸、接收、播放，系統(tǒng)能夠根據(jù)網(wǎng)絡(luò)帶寬的變化對(duì)接收媒體段的碼率做出合適的調(diào)整，在不同網(wǎng)絡(luò)環(huán)境下能夠流暢的觀(guān)看視頻。 本文在研究H.265視頻壓縮技術(shù)和MPEG-DASH流媒體技術(shù)的基礎(chǔ)上，，利用PSNR和SSIM對(duì)H.265的視頻客觀(guān)質(zhì)量進(jìn)行了評(píng)價(jià)，并提出了基于視頻質(zhì)量的自適應(yīng)算法。最后基于libdash庫(kù)實(shí)現(xiàn)了完整的流媒體系統(tǒng)，并在不同場(chǎng)景下測(cè)試了系統(tǒng)的功能。本文所實(shí)現(xiàn)的系統(tǒng)為以后流媒體系統(tǒng)開(kāi)發(fā)提供了指導(dǎo)。
[Abstract]:With the continuous construction of the network and the increase of the number of high-definition video, the proportion of streaming media service in the whole network traffic is increasing, which makes the streaming media technology constantly studied at home and abroad. Aiming at the shortcomings of the current streaming media solution, this paper proposes a streaming media solution based on H.265 and MPEG-DASH. In this paper, the key technology of H.265 video compression standard is analyzed, and the content of H.265 video quality evaluation is discussed. Two parameters, PSNR and SSIM, are selected to evaluate H.265 video quality objectively. Secondly, the MPEG-DASH correlation technology is studied, the current adaptive algorithm is analyzed, and the video quality adaptive algorithm based on H.265 video quality evaluation is proposed. Based on the bandwidth-based adaptive algorithm, the video quality and the usage of receive buffers and their variations are taken into account in the process of bitstream handover. The video quality can be used to improve the problem of excessive picture change in the process of bitstream handover. For the consideration of buffer, based on the previous algorithm, the switching problem of high or too low buffer is considered. The adaptive algorithm based on video quality is implemented by using matlab platform, and the function of the algorithm is verified by simulation. The performance of the improved algorithm is superior to that of the adaptive algorithm based on bandwidth when the media in buffer is less, too much, and the bandwidth of the network changes greatly. At the end of this paper, we combine H.265 and MPEG-DASH to realize a streaming media system based on libdash. The system uses MPEG-DASH as the transmission protocol, H.265 is used to compress the video on the server, and the client uses multithreading technology based on libdash library. An adaptive algorithm based on bandwidth is implemented. The system is tested in six scenarios in LAN and WAN. The test results show that the system can transmit, receive and play the dash stream compressed by H.265. The system can adjust the bit rate of received media segment according to the change of network bandwidth, and can watch video smoothly in different network environment. Based on the research of H.265 video compression technology and MPEG-DASH streaming media technology, the objective quality of H.265 video is evaluated by PSNR and SSIM, and an adaptive algorithm based on video quality is proposed. Finally, a complete streaming media system based on libdash library is implemented, and the functions of the system are tested in different scenarios. The system realized in this paper provides guidance for the future development of streaming media system.
【學(xué)位授予單位】：武漢理工大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2014
【分類(lèi)號(hào)】：TN919.8

【參考文獻(xiàn)】

相關(guān)博士學(xué)位論文 前1條

1 元輝;新一代視頻編碼技術(shù)研究[D];西安電子科技大學(xué);2011年

本文編號(hào)：1905222