AVS2视频解码器在TMS320C6678上的优化和实现

发布时间：2018-01-07 07:28

本文关键词：AVS2视频解码器在TMS320C6678上的优化和实现　出处：《天津大学》2014年硕士论文　论文类型：学位论文

更多相关文章： AVS2 解码 TMS320C6678 移植优化

【摘要】：现在国际上主流的视频标准主要有H.26x系列和MPEG系列。其中,H.264标准的复杂度非常高、专利费也很昂贵。AVS2是我国新一代具有自主知识产权的音视频编解码标准,在性能方面与H.264相当,但在复杂度方面却远远低于H.264,而且其专利费比H.264便宜很多。目前,AVS2视频解码器在软件方面和硬件方面的发展不够成熟,离实现产业化的目标还有很长的路要走。DSP以其数字信号处理方面的优势,特别适合于音视频编解码应用。基于以上的背景,本文选取TMS320C6678作为硬件平台,提出了AVS2解码器在TMS320C6678上的优化和实现方案。论文的主要研究内容如下:首先,阐述了视频压缩技术的研究现状、课题的研究背景及研究意义。详细介绍了国内外视频编码标准的发展历程,分析了视频解码器的关键技术以及解码器的体系结构,详细介绍了本文使用的硬件平台和开发工具CCS,以及DSP软件开发与优化流程。然后,阐述了将AVS2视频解码器移植到TMS320C6678平台的过程。由于TMS320C6678平台的开发环境和资源与PC平台的并不相同,移植的过程中,分别从运行支持库、内存分配、变量定义等方面进行调整,最终AVS2视频解码器可以正常运行在TMS320C6678平台上。最后,对该解码器进行优化。在研究AVS2视频编解码标准的基础上研究了AVS2视频解码器的关键技术,例如:熵解码、反量化和反变换、帧内和帧间预测、环路滤波等。接着从DSP平台和C语言级两个方面对其进行了优化。最后,利用Intel VTune分析工具对视频解码器的瓶颈和耗时情况进行分析,并根据分析结果对视频解码器的关键模块进行了重点优化。测试结果表明,优化后的视频解码速度得到了很大提高,1280*720分辨率的视频能够达到每秒30帧的解码要求。
[Abstract]:At present, the main international video standards are H.26x series and MPEG series, among which the complexity of H.264 standard is very high. The patent fee is also very expensive. AVS2 is a new generation of audio and video coding and decoding standards with independent intellectual property rights in China, which is comparable to H. 264 in performance, but far lower in complexity than H. 264. Moreover, the patent fee is much cheaper than H.264. At present, AVS2 video decoder is not mature in software and hardware. There is still a long way to go to achieve the goal of industrialization. DSP has its advantages in digital signal processing, especially for audio and video coding and decoding applications. Based on the above background. In this paper, TMS320C6678 is selected as the hardware platform, and the optimization and implementation of AVS2 decoder on TMS320C6678 is proposed. The main contents of this paper are as follows: first. The research status, research background and significance of video compression technology are described. The development of video coding standards at home and abroad is introduced in detail. This paper analyzes the key technology of video decoder and the architecture of the decoder, introduces the hardware platform and development tool CCS, as well as the DSP software development and optimization process in detail. This paper describes the process of transplanting AVS2 video decoder to TMS320C6678 platform because the development environment and resources of TMS320C6678 platform are different from that of PC platform. In the process of transplanting, we adjust the running support library, memory allocation, variable definition and so on. Finally, the AVS2 video decoder can run normally on the TMS320C6678 platform. Finally. The key techniques of AVS2 video decoder, such as entropy decoding, inverse quantization and inverse transform, intra-frame and inter-frame prediction, are studied based on the research of the standard of AVS2 video codec. Loop filter and so on. Then it is optimized from DSP platform and C language level. Finally, the bottleneck and time consuming of video decoder are analyzed by using Intel VTune analysis tool. The key modules of the video decoder are optimized according to the analysis results. The test results show that the optimized video decoding speed has been greatly improved. Video with resolution 1280 / 720 can achieve decoding requirements of 30 frames per second.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN919.81

【相似文献】