本文選題：H.264視頻編碼　切入點：資源分配　出處：《哈爾濱工業(yè)大學》2014年碩士論文　論文類型：學位論文

【摘要】：隨著移動互聯(lián)網(wǎng)的快速發(fā)展，數(shù)據(jù)通信尤其是視頻通信得到了比以往更廣泛的應用。H.264是面向未來IP和無線網(wǎng)絡傳輸環(huán)境提出的新一代視頻編碼標準，在壓縮效率和網(wǎng)絡適應性等方面性能顯著提高，但是H.264高效的編碼性能是以其巨大的計算復雜度為代價的。同時，H.264的碼率控制因為其采用的率失真優(yōu)化技術而變得更加復雜。 研究表明，不同視頻場景對人類視覺系統(tǒng)（Human Visual System, HVS）的重要性是不同的，然而傳統(tǒng)的視頻編碼算法并沒有充分考慮HVS對視頻場景感知的多樣性。因此，如何利用HVS特性，對H.264編碼的計算資源和比特資源進行分配，以改善H.264的編碼性能，具有重要的理論意義和應用價值。本文正是在這種背景下，對基于感興趣區(qū)域的視頻編碼資源分配算法做出一定貢獻： 首先，對視頻編碼的研究發(fā)展歷程做了簡要介紹，具體介紹了H.264視頻編碼標準，對其編碼框架、關鍵的技術特征和碼流結構都進行了理論分析。同時，詳細介紹了H.264所采用的三層控制結構的碼率控制器，并詳述了現(xiàn)有的碼率控制技術缺陷和基于感興趣區(qū)域的比特資源分配的可能性。 其次，提出了基于感興趣區(qū)域的視頻編碼計算資源分配算法。在滿足感興趣區(qū)域的視頻編碼質量前提下，對視頻編碼計算資源進行優(yōu)化分配，以達到降低編碼復雜度，使其適合實時傳輸處理的目的。對比分析該算法相對于H.264標準JM代碼，在不降低感興趣區(qū)域的視頻質量情況下，計算節(jié)省編碼時間的比例，編碼效率更高。 最后，提出了基于感興趣區(qū)域的視頻編碼比特資源分配算法。該算法在低碼率視頻編碼傳輸比特資源受限的情況下，通過控制不同區(qū)域的量化參數(shù)，能夠將更多的比特資源分配給人眼感興趣區(qū)域，并通過感興趣區(qū)域分級，將不同區(qū)域視頻質量平滑過渡。該算法在變碼率情況下仍然適用，，實驗結果表明，該算法能夠實現(xiàn)碼率控制，并提高感興趣區(qū)域的PSNR值，節(jié)省比特資源。
[Abstract]:With the rapid development of mobile Internet, data communication, especially video communication, has been more widely used than ever before. H.264 is a new generation video coding standard for future IP and wireless network transmission environment. The compression efficiency and network adaptability are improved significantly, but the efficient coding performance of H.264 is at the expense of its huge computational complexity. Meanwhile, the rate control of H.264 becomes more complex because of the rate-distortion optimization technology used in H. 264. The research shows that the importance of different video scenes to human visual system is different. However, the traditional video coding algorithms do not fully consider the diversity of HVS's perception of video scenes. Therefore, how to make use of the HVS characteristics, It is of great theoretical significance and practical value to allocate computing resources and bit resources of H.264 coding in order to improve the coding performance of H.264. It makes some contributions to the algorithm of resource allocation for video coding based on region of interest (ROI). Firstly, the research and development of video coding are briefly introduced, and the H.264 video coding standard is introduced in detail. The coding framework, key technical features and code stream structure are analyzed theoretically. This paper introduces the rate controller of the three-layer control structure used in H.264 in detail, and describes the defects of the existing rate control technology and the possibility of bit resource allocation based on the region of interest. Secondly, this paper proposes a video coding computing resource allocation algorithm based on the region of interest. Under the premise of satisfying the video coding quality of the region of interest, the video coding computing resources are allocated optimally to reduce the coding complexity. Compared with the standard JM code of H.264, the proposed algorithm is more efficient than the JM code of H.264, and the ratio of saving coding time is calculated without reducing the video quality of the region of interest. Finally, a video coding bit resource allocation algorithm based on the region of interest is proposed, which can control the quantization parameters of different regions in the case of low bit-rate video coding transmission bit resource constraints. More bit resources can be allocated to the region of interest, and the video quality of different regions can be smoothly transitioned through the classification of the region of interest. The algorithm is still applicable in the case of variable bit rate. Experimental results show that, The algorithm can achieve bit rate control, improve the PSNR value of the region of interest, and save bit resources.
【學位授予單位】：哈爾濱工業(yè)大學
【學位級別】：碩士
【學位授予年份】：2014
【分類號】：TN919.81

【參考文獻】

相關期刊論文 前5條

1 王曉龍;崔慧娟;唐昆;;H.264/AVC中基于感興趣區(qū)域的計算資源的優(yōu)化分配算法[J];清華大學學報(自然科學版);2011年11期

2 李國輝;梅魁志;袁澤劍;;基于膚色模型和貝葉斯判別的人臉檢測[J];計算機應用;2006年08期

3 姜軍,張桂林;一種基于知識的快速人臉檢測方法[J];中國圖象圖形學報;2002年01期

4 盧春雨,張長水,聞芳,閻平凡;基于區(qū)域特征的快速人臉檢測法[J];清華大學學報(自然科學版);1999年01期

5 王見軍;畢厚杰;;低比特率通信的H.263視頻編解碼[J];中興新通訊;1996年02期

本文編號：1563025