【作者】 柏海涛；

【导师】 张建州；

【作者基本信息】 四川大学 ， 计算机应用技术， 2005， 硕士

【摘要】 随着人们对视频和音频信息的需求愈来愈强烈,追求远距离视音频同步交互成为新的时尚。近些年来,依托计算机技术、通信技术和网络技术的发展,集音频、视频、图像、文字、数据为一体的多媒体信息,使越来越多的人开始通过互联网享受到网上生活、远程医疗、远程通讯的乐趣,缩短了时区和地域的距离。在这之中,数字电视(包括HDTV)、DVD 压缩与存储、视频会议和以多媒体应用为主要特点的3G 都是研究热点,也是多媒体技术的一个发展方向和趋势。而这其中的一个关键技术就是视频压缩技术,促使了许多视频编码标准的产生。H.264/AVC 是由ITU(国际电信联盟)的VCEG 和ISO/IEC 的MPEG 两大组织联合组成的JVT 共同制定的一项新的视频压缩技术标准,它使得运动图像压缩技术上升到了一个更高的阶段。在较低带宽上提供高质量的图像传输是H.264/AVC 的应用亮点,在同样的视觉质量前提下,比H.263 和MPEG-4 节约了50%的码率,而且对IP 和无线网络传输具有更好的支持功能。但H.264 获得优越性能的代价是计算复杂度的增加,这也是H.264 大规模应用的主要障碍。针对这一现状,业内也在做大量的研究,力图降低其计算复杂度和提高运行效率。本文的工作主要集中以下几个方面: 1. 仔细研究了H.264 的标准,并与其他标准进行了较为全面的对比,从标准的新特点中确定了研究的重点与方向。2. 通过对JM 参考模型代码的研究,找出其中的计算瓶颈和耗时、低效的模块,并设计了一个新的模型。3. 针对计算量大、复杂度高的运动估计和补偿模块,用快速算法和简化算法减少计算量,提高运算效率。4. 针对计算量大、频繁调用,但是规则性强的运动估计插值和整数变换更多还原

【Abstract】 With the explosive growth of the demand for video and audio information, the pursuing of interactivity of video and audio over long-distance has been fashioned. In recent years, based on the develop of computer science, communication and networking, multimedia information which integrated audio, video, images, texts and data, has facilitated more and more people by bringing the joys of cyber-life, tele-medical and telecommunications, which shortened the distance over time and space. Among this, digital TV(including HDTV), compressing and storage of DVDs, video conference and 3G is the focus of research and one of the trend of multimedia technologies. The key technology among these applications is video compressing, which brings many coding standards out. H.264/AVC is newest video coding standard of the ITU-T Video Coding Experts Group (VCEG) and the ISO/IEC Moving Picture Experts Group (MPEG), which greatly enhance compressing performance and provide a “network-friendly”video representation addressing “conversational”( video telephony) and “non -conversational”(storage, broadcast, or streaming) applications. But the enhancement of compressing is based on rapid increasing of compute complexity, which has been the main hindrance of massive application. Aiming at the status quo, many researches have been done and being done to decrease compute complexity and together improve running efficiency. Our work mainly focused on the following: 1. Analyzed the standard in detail and compared with other standards comprehensively, to establish the emphases and orientation of further study. 2. Based on the study of JM, we found out the computing bottleneck and the time-consuming, inefficient modules and set up a new module. 3. Aiming at the modules that contain massive computation and complex, we use fast or predigest algorithm to cut down the computation amount to enhance the efficiency. 4. Aiming at the modules that are frequently called but well formed, we use MMX, SSE or SSE2 to optimize, which would bring a higher performance. 5. Some analysis was given on the relation of rate control and picture quality, and we point out the parameters of common applications and give perspectives for further development.更多还原