【作者】 王立；

【导师】 滕建辅； 李桂苓；

【作者基本信息】 天津大学 ， 信号与信息处理， 2007， 博士

【摘要】 随着通信、计算机、数字音视频和网络技术的发展以及多种终端的大量应用,人们已经不再满足于单纯传递文字和声音,而希望能高速传送数据量更大的视频信息。在这种需求推动下,近些年来数字视频压缩技术获得了很大发展,面向不同应用,相继出现多种音视频编码标准,并随之产生了不同格式视频资源共享问题,进而催生了转换编码技术。由于新视频编码标准的出现,视频转换编码技术的应用范围越来越广。本文首先对转换编码技术进行了分析和论述,对一些常用的转换编码器结构进行了系统分析。之后对目前主流的两个音视频编码标准:MPEG-2和H.264/AVC进行了较深入地研究,指出了MPEG-2到H.264/AVC转换编码技术的难点。本文第三章提出了MPEG-2到H.264/AVC转换编码中模式判断快速算法,包括帧内预测和前向帧间预测两部分。根据MPEG-2中视频DCT系数的特点和宏块的编码模式,在DCT域确定H.264/AVC中帧内宏块的编码模式,通过跳过一些不必要的宏块模式判断等,节省大量的计算量,提高转换编码效率。第四章对MPEG-2到H.264/AVC转换编码中的运动矢量优化算法进行了研究。通过重用MPEG-2各帧间预测宏块的运动矢量,提出了不同的运动矢量优化算法,并充分利用H.264/AVC运动估计新技术,使运动矢量精度提高到1/4像素,并有效地降低运动估计计算量,提高转换编码运算速度。第五章研究了转换编码技术在数字电视中的应用,包括台标插入系统和传送流无缝拼接。在台标插入系统中,利用同类转换编码技术实现了视频信息的插入。在此基础上,本文提出了精确控制码率的算法,保证基本流输出码率不变。在传送流无缝拼接的研究中提出了两种算法:基于系统层的算法计算量小,运算速度快;基于基本层的算法通过改变视频帧编码类型实现,解决了精确到帧的无缝拼接问题。第六章遵照MPEG-2系统层的系统目标解码器(T-STD)模型,提出了多路复用算法,高效实现多路音视频数据流复用。在多路复用算法基础上,结合数字电视台标插入系统需要,研究了传送流二次复用算法,在保证输出码流正确解码的同时,有效的降低了复用计算量。本文在系统研究数字视频转换编码技术的基础上,对MPEG-2到H.264/AVC异类视频转换编码算法以及将同类转换编码技术用于数字电视系统的研究成果有很高实用价值,有些已投入实际应用,取得了良好效果。更多还原

【Abstract】 With the development of the Internet technology, people want to transmit not only word and sound, but also video information. The digital video technology has developed quickly in order to transmit video in high quality. Because many kinds of video coding standards are used, it is a problem to transform different format of the program. Video transcoding technology is the key method for solving such problems.Video transcoding has been applied in new area for new video coding standard. In this disertation, the overview of transcoding is presented first. Some transcoding structures are described and analyzed. Then two video standards, that is, MPEG-2 and H.264/AVC are studied in chapter 2. The difficulties in transcoding from MPEG-2 to H.264/AVC are pointed out and mode decision and motion estimation are investigated. Some algorithms for transcoding are proposed in Chapter 3 and 4. The experiment results show that the proposed algorithms have little loss in video quality and remarkablely increasing in transcoding speed. It is suit for transcoding applications.In chapter 5, two transcoding applications in DTV are dealt with. One is logo inserter and the other is seamless splicing for transporting stream. The rate control algorithm of logo inserter controlls the output element stream accurately and ensures the output frame length close to input frame length. Two algorithms on seamless splicing are provided. One is based on system layer. The other is based on video element stream. They can both splice two transporting streams seamlessly. The first one possesses less calculation and higher speed characteristics. The second one can splice more accurately.In chapter 6, multiplexing and re-multiplexing technology are examed. The multiplexing algorithm is in accordance with T-STD. The efficiency is improved and the bandwidth is saved. The re-multiplexing algorithm is researched for logo inserter. It is based on some special structure of logo inserter. So the calculation is decreased and operation speed is high. As a result, less system resource is used.This disertation is mainly concerned with the transcoding technologies from MPEG-2 to H.264/AVC and transcoding applications in DTV. Some algorithms are proposed and verified by experiments. The algorithms are efficiency and some of them have been used in practice.更多还原