H.264优化算法及其在嵌入式WinCE6.0中的应用

发布时间：2018-03-30 11:38

本文选题：H.264算法优化　切入点：帧内预测　出处：《兰州理工大学》2012年硕士论文

【摘要】：近些年来视频解码技术飞速发展，低分辨率、低质量的多媒体视频已不能满足人们的需求。近些年来随着3G技术以及智能手机的广泛应用，人们对手持多媒体视频传输的需求越来越迫切，但是移动终端或者无线技术传输的过程中对多媒体信息传输，尤其是高质量的视频多媒体传输对系统硬件以及网络带宽往往具有很高的要求。所以，视频的传输需要进行压缩处理。视频编解码技术是实现手持终端视频无线传输的重要核心技术。H.264编解码技术已经广泛应用在PC机为主的家用平台，成为下一代视频压缩技术强有力的竞争者。但在手持移动终端设备当中，H.264由于其编码较为复杂并且对硬件处理性能要求较高等原因还未大量使用。如何能够将已经非常优秀的H.264编解码技术应用在手持移动终端，其关键是在尽量不改变视频质量的情况下相应的提高编码效率，降低算法的复杂度，在硬件环境有限的情况下满足H.264编解码正常运行并达到良好效果，所以对其算法优化具有重要的研究意义。根据上述特点及H.264算法中需要解决的问题，本文从以下几点对H.264算法进行研究与优化。 1、深入学习H.264算法的几种关键技术，了解这些关键技术国内外研究现状，通过对比其它解码器，解释了H.264为何优于其它算法，为以下研究打下基础。 2、由于静止画面中其像素之间相关性非常强，通过对帧内预测算法深入研究，本文提出了一种基于阈值法与方向梯度法相结合的快速选择预测模式算法。该算法首先将用设置阈值的方式将算法简化，如果满足阈值设置条件直接进入DC模式，如果不满足条件进入简化后的方向预测算法。仿真结果显示，利用该算法能够在基本不改变视频质量的情况下降低60%以上的运算时间，具有一定的实用性。 3、上述方法能够很解决静止画面的压缩效率，但对运动较为强烈的画面容易产生图像失真和较强的飘移效应。对于运动剧烈的画面本文对运动补偿算法分析并优化该算法，提出一种基于1/4像素法的快速算法。在实验分析计算的基础上得到像素中需要大量计算的部分并引入用于像素插值滤波器进行滤波，通过此算法能够降低运算的复杂度，减少运算时间，，在实验结果中证明该算法可以有效地降低剧烈运动画面中总的计算复杂的解码过程，而不会带来明显的图像质量损失和严重的漂移效应。 4、对帧内预测算法以及运动补偿算法优化后的算法移植入嵌入式系统当中，论文选择基于ARM11的S3C6410开发板。将WinCE6.0移植入开发板中，并通过制作简单播放器来测试视频优化后的效果。
[Abstract]:With the rapid development of video decoding technology in recent years, the low resolution and low quality multimedia video has not been able to meet the needs of people. In recent years, with the wide application of 3G technology and smart phone, The demand for hand-held multimedia video transmission is becoming more and more urgent, but in the process of mobile terminal or wireless technology transmission of multimedia information, In particular, high quality video multimedia transmission often requires high system hardware and network bandwidth. Video coding and decoding technology is an important core technology for wireless video transmission in handheld terminal. H.264 coding and decoding technology has been widely used in PC based home platform. It has become a strong competitor in the next generation video compression technology. However, H.264 has not been widely used in handheld mobile terminal devices due to its complex encoding and high performance requirements for hardware processing. It has been applied to handheld mobile terminal with excellent H.264 coding and decoding technology. The key is to improve the coding efficiency and reduce the complexity of the algorithm without changing the video quality as much as possible, and to meet the H.264 coding and decoding normal operation and achieve good results under the limited hardware environment. Therefore, it is of great significance to study its algorithm optimization. According to the above characteristics and the problems that need to be solved in H.264 algorithm, this paper studies and optimizes H. 264 algorithm from the following points. 1. Study several key technologies of H.264 algorithm, understand the research status of these key technologies at home and abroad, explain why H.264 is superior to other algorithms by comparing other decoders, and lay the foundation for the following research. 2. Because of the strong correlation between the pixels in the still picture, the intra prediction algorithm is studied deeply. In this paper, a fast predictive mode selection algorithm based on the combination of the threshold method and the directional gradient method is proposed. The algorithm will be simplified by setting the threshold, if the threshold setting condition is satisfied and the DC mode is entered directly. The simulation results show that the algorithm can reduce the computation time by more than 60% without changing the video quality, and it has some practicability. 3. The above method can solve the compression efficiency of still picture very well, but it is easy to produce image distortion and strong drift effect for the picture with strong motion. In this paper, the motion compensation algorithm is analyzed and optimized for the picture with violent motion. In this paper, a fast algorithm based on 1 / 4 pixel method is proposed. Based on the experimental analysis and calculation, the part of pixel which needs a lot of computation is obtained and the filter is used to filter the pixel interpolation filter. Through this algorithm, the computational complexity can be reduced. The experimental results show that the proposed algorithm can effectively reduce the total computational complexity of the decoding process in the violent motion picture without obvious loss of image quality and serious drift effect. 4. For the intra-frame prediction algorithm and the optimized algorithm of motion compensation algorithm transplanted into the embedded system, the S3C6410 development board based on ARM11 is selected in this paper. The WinCE6.0 is transplanted into the development board. And through the production of simple players to test the effect of video optimization.
【学位授予单位】：兰州理工大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP368.1;TN919.81

【参考文献】