【作者】 陈强；

【导师】 古辉；

【作者基本信息】 浙江工业大学 ， 计算机应用技术， 2013， 硕士

【摘要】 H.264/AVC视频编码标准,具有高效的压缩率和较好的网络适应性,广泛应用于网络流媒体(如YouTube、iTunes Store)、高清电视(如ATSC、DVB-T)、卫星(DVB-S、DVB-S2)等领域。由于它采用率失真优化技术(RDO, Rate Distortion Optimization),在提高编码效率的同时也大大增加了计算复杂度,在一定程度上限制了其应用。为了降低H.264视频数据编码算法的计算复杂度,兼顾和提高编码效率,较好的体现实时性,本文利用图像纹理特性以及子块之间的相关性,提出一种快速帧内预测模式选择算法,通过JM8.6测试模型对其编码时间、码率以及图像质量进行验证。此外,通过TI公司的TMS320DM365处理器设计一款校车远程监控系统对H.264进行应用。本文的主要工作如下：1.介绍了本文的研究目的和意义、视频编码标准的发展历程,阐述了H.264视频编码技术,着重对帧内预测的原理、率失真优化技术、预测模式全搜索算法进行研究。2.提出一种快速帧内预测模式选择算法。①根据图像的纹理方向,减少预测模式的数量；②根据色度块与亮度块预测模式的相对独立性,减少预测模式的重复计算；③根据子块之间模式的相关性,进一步减少RDO的计算。实验结果显示,在图像质量基本保持不变的情况下,编码时间平均减少65%左右。3.设计校车远程监控系统原型,并将H.264视频数据编码进行应用。以TMS320DM365为核心处理器,结合视频采集模块、电源模块等构建校车远程监控系统模型。实验结果表明,基本实现了远程实时监控的目标。更多还原

【Abstract】 H.264/AVC-the ITU-T VCEG and the ISO/IEC MPEG jointly developed a new generation video coding standard, which has higher compression efficiency and improved network adaptation, used in the field of high-performance video services widely. It adopts RDO(Rate Distortion Optimization) technique, which improves the encoding effective, but increases computational complexity. In order to reduce the computational complexity of the H.264video coding algorithm, improve the coding efficiency, and achieve the goal of real-time performance, this paper adopts the texture features and the correlation between the sub-block to present a fast intra prediction mode selection algorithm. What’s more, we use the JM8.6standard test models to verify the encoding time、bit rate and image quality. In addition, this paper designs a school bus remote monitoring system based on the TMS320DM365processor from TI company for the Ⅱ.264video coding application. The following are the main contributions of this paper.1. This paper describes the purpose and significance of the research, and the video coding standard develpment process, elaborates H.264video coding technology, focuses on the intra prediction theory, RDO techniques, full search algorithm of the prediction mode.2. A fast intra prediction mode selection algorithm is proposed. First, according to the texture direction, reduce the number of prediction modes. Second, according to the relative independence of chrominance and luminance block prediction mode, reduce double counting of the prediction mode. Third, according to sub-block mode correlation, reduce the RDO calculation further. Test shows that the algorithm reduce the encoding time about65%with negligible loss of PSNR.3. A school bus remote monitoring system based on the TMS320DM365processor for the H.264video coding application is proposed. With the TMS320DM365processor, combined with video capture module, power supply module to build the school bus remote monitoring system models. Test shows that this model achieves the goal of real-time and remote performance basicly.更多还原