HEVC像素梯度帧内预测算法设计与实现

发布时间：2018-02-20 23:31

  本文关键词： HEVC 像素梯度 帧内预测　出处：《电子科技大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着互联网技术的发展,越来越多的人使用智能手机、平板电脑观看在线高清视频。同时,超高清视频技术正逐步兴起。在线高清视频流给网络设备造成了沉重的负担,而目前主流的H.264视频编码标准对高清、超高清的压缩性能较低。为进一步降低视频编码后码率,提高对高清、超高清视频的压缩性能,JCT-VC组织制定了新一代视频编码标准HEVC。HEVC基于H.264中的混合视频编码框架,引入了灵活单元划分、更加精细的帧内预测、环内去块滤波等技术。但是灵活的单元划分和更加精细的帧内预测技术给HEVC帧内编码带来了更高的计算复杂度,因此如何有效减少帧内编码时间,已经成为HEVC算法研究中的一个研究热点。文中利用像素点的梯度信息,改进了HEVC帧内预测算法,并使用C++语言进行算法的实现,最后给出了算法性能分析结果。具体内容如下:首先,研究了HEVC基本编码结构及其帧内预测关键技术,并同H.264帧内编码技术进行了对比分析。其次,研究分析H.264、HEVC帧内预测算法,重点分析了H.264中基于图像纹理特征的帧内预测算法和HEVC粗选择帧内预测算法。在此基础上,使用像素梯度信息,改进了粗选择帧内预测算法,并使用C++语言实现该算法。最后,使用HEVC测试视频,在全帧内、低复杂度编码模式下,对像素梯度帧内预测算法和粗选择算法进行对比测试。测试结果表明,像素梯度帧内预测算法在相同图像质量的情况下,码率平均增加1.62%,编码时间平均减少22.50%,证明算法有效可行。像素梯度帧内预测算法对粗选择帧内预测算法进行了改进,有效减少了帧内预测编码时间,具有一定的理论意义与实践价值。
[Abstract]:With the development of Internet technology, more and more people use smart phones and tablets to watch online HD video. Meanwhile, UHD video technology is gradually rising. However, the current mainstream H.264 video coding standard has low compression performance for high-definition and ultra-high-definition video. In order to further reduce the rate of video coding and improve the high-definition, High definition Video Compression performance JCT-VC organizes and formulates a new generation of video coding standard HEVC.HEVC based on H.264 hybrid video coding framework, introduces flexible unit partitioning, more refined intra prediction, However, flexible unit partition and finer intra prediction technology bring higher computational complexity to HEVC intra coding, so how to reduce the intra coding time effectively. In this paper, the gradient information of pixel points is used to improve the HEVC intra prediction algorithm, and C language is used to implement the algorithm. Finally, the performance analysis results of the algorithm are given. The main contents are as follows: firstly, the basic coding structure of HEVC and the key techniques of intra prediction are studied, and compared with H.264 intra coding. Secondly, the intra prediction algorithm of H.264 and HEVC is studied and analyzed. The intra prediction algorithm based on image texture feature and the HEVC coarse selection intra prediction algorithm in H.264 are analyzed. Based on this, the coarse selection intra prediction algorithm is improved by using pixel gradient information, and the algorithm is implemented in C language. The HEVC test video is used to compare the pixel gradient intra prediction algorithm with the coarse selection algorithm in the full frame and low complexity coding mode. The test results show that the pixel gradient intra prediction algorithm is of the same image quality. The average bit rate increases by 1.62bit rate and the coding time decreases by 22.50. it is proved that the algorithm is effective and feasible. The pixel gradient intra prediction algorithm improves the coarse selection intra prediction algorithm and effectively reduces the intra prediction coding time. It has certain theoretical significance and practical value.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN919.81
，

本文编号：1520343