【作者】 崔玉斌；

【导师】 蔡安妮；

【作者基本信息】 北京邮电大学 ， 通信与信息系统， 2008， 博士

【摘要】 在过去十年中,无线通信和多媒体通信技术取得了飞速发展,并获得了巨大的商业成功。无线通信和多媒体技术的融合正在加速进行,在视频压缩算法、视频传输网络架构与协议、数字信号处理芯片等方面的研究成果不断涌现。面向无线通信的视频图像编解码与传输技术已成为当今信息技术领域的前沿课题之一。H.264/AVC是由国际标准化组织/国际电工委员会(ISO/IEC)中的运动图像专家组(MPEG)和国际电信联盟(ITU-T)中的视频编码专家组(VCEG)结合起来组成的联合视频工作组(JVT)制订的新一代国际视频编码标准,采用了近几年来视频编码方面的先进技术,以较高编码效率和网络友好性在各种视频服务中被广泛地应用。本文针对基于H.264的无线视频通信系统中的关键技术展开了深入细致地研究。具体包括:视频信源编码中的宏块模式选择快速算法、速率控制算法和容错编码方法;信道编码中的不均衡差错保护;和基于H.264的扩展空域可分层编码中增强层宏块的快速模式选择和细粒度可分层编码(FGS)中的抗漂移差错方法。致力于解决H.264编码器计算复杂度高而移动终端计算能力较弱的问题,H.264视频码流在易错的无线网络中鲁棒传输的相关问题,和如何在目前广泛存在的异构网络中开展H.264视频应用的问题。本文的主要内容和创新之处为:首先,介绍了基于H.264的无线视频通信系统,并对H.264码流中宏块/子宏块的编码模式和编码过程中的重要参数进行了统计分析,为后续提出的各种算法提供了依据。其次,分别提出了降低帧内(I)帧编码计算复杂度的帧内预测快速算法和降低预测(P)、双向预测(B)帧宏块模式选择计算复杂度的快速算法。再次,采用信源信道联合编码的思想,提出了一种把速率控制与容错编码和不均衡差错保护有机地结合起来的无线视频通信解决方案。最后,使用基本层局部宏块的运动和纹理特性,分别提出了降低拓展空域可分层(ESS)编码中增强层宏块模式选择的计算复杂度和对自适应参考帧FGS(AR-FGS)编码中采用的泄漏系数α进行局部自适应调整的方法。更多还原

【Abstract】 In the past decade,a rapid development of wireless communication and multimedia communication techniques has been achieved,and it resulted in great commercial success.The merge of wireless communication and multimedia techniques is accelerating recently and productions emerge continuously such as video transmission network protocols,powerful digital video signal processing chips as well as efficient video compression algorithms.The research on the video codec and video stream transmission in wireless environment has become an active task in information science and technology field.H.264/AVC is the newest international video coding standard approved by the joint video team(JVT) organization which is formed by ITU-T VCEG(Video Coding Experts Group) and ISO/IEC MPEG (Moving Picture Experts Group).H.264 adopts a number of advanced video coding technologies and has been used in all sorts of video applications because of its high coding efficiency and network friendliness.This dissertation has conducted the research focusing on the key technologies in H.264 based wireless video communication system.The contents of the task are mainly composed of fast macroblock(MB) mode selection,rate control and error resilience coding in video encoder; unequal error protection(UEP) in channel encoder;fast mode derivation of the enhancement layer MB in extended spaticial scalability(ESS) and drift resistance in fine granularity scalability(FGS) in H.264/AVC scalable video coding(SVC) extension.Our works try to decrease the computational complexity of H.264 encoder to adapt the calculation ability of mobile terminals,solve the problems of robust transmission of H.264 video over error-prone wireless networks as well as enable flexible applications over heterogeneity networks.The contents and innovations of this dissertation are as follows. First,we describe the architecture of H.264/AVC based wireless video communication system,process statistical analyses on MB/sub_MB mode in H.264 streams as well as some important encoding parameters and draw some conclusions which are foundations of our following proposed algorithms.Next,we propose a fast intra prediction algorithm (?) reduce the computational complexity of encoding intra(I) frame, (?)nd a fast MB mode decision algorithm to reduce the complexity of (?)coding predict(P) and bi-predict(B) frames.Again,we propose a solution for wireless video communication which adopts the rate control associate with the error resilience encoding and UEP motivated by the joint source and channel coding(JSCC).Finally,in SVC extension (?) H.264/AVC,we proposed an efficient method to reduce the computational complexity of the mode derivation process of the hancement-layer MBs in ESS as well as a local adaptation of leak factorαused in the leaky prediction scheme in adaptive reference FGS (AR-FGS) respectively,which both depend on the motion and texture characteristics of the co-located local MBs in the base layer.更多还原