基于解码的空间可伸缩AVS视频编码
发布时间：2018-12-31 21:27
【摘要】：随着视频压缩编码技术和网络技术的不断发展,基于网络传输的视频得到了广泛的应用。长期以来,我国在数字音视频产业发展方面一直受制于国外的视频压缩编码标准,每年都要支付高额的专利费用。为此我国成立了AVS工作组,制定了具有我国自主知识产权的数字音视频编解码标准AVS。AVS标准具有编解码效率高、编解码算法复杂度低和软硬件实现成本低等优势,未来必然能在数字音视频领域得到广泛的应用。 基于网络传输的视频编码与面向存储的视频编码存在着很大的差异,它必须考虑到网络的异构性以及终端设备的多样性等因素,因此编码器所编出的码流应在空间、时间、质量等方面可伸缩,这样的码流才具有很好的适应能力,可伸缩视频编码(Scalable Video Coding, SVC)技术就在这样的环境下应运而生。SVC实现了只对视频序列进行一次编码,接收终端就可根据网络连接情况和自身的显示尺寸,对编码码流部分解码或者部分传输,从而保持带宽均衡和清晰流畅的多画面显示。毫无疑问,对可伸缩视频编解码技术的深入研究具有极其重要的意义。 本课题的设计目标是实现可伸缩视频编解码系统,故首要工作是实现一个优化的AVS解码器,接着修改AVS编解码算法,最终实现AVS空间可伸缩编码以及码流的正确解码。主要研究的内容及成果如下： ●在对AVS编解码原理及官方发布的参考解码软件深入研究的基础上,提出了新的解码流程。首先在PC机上用C代码实现了新的AVS解码器,并对该解码器进行C代码级别的优化。重点采用SIMD技术中的MMX指令集和SSE2指令集对新解码器中运算量集中且耗时较多的代码段进行改写。经过优化后的新解码器实现了D1格式图像基于PC平台的实时解码。 ●通过对可伸缩视频编码原理和关键技术的研究,提出了一个AVS空间可伸缩视频编解码方案,与SVC相比较为简单。首先按照该方案对编码器进行相关代码的修改,修改之后的编码器所编出的码流可以提供多种分辨率的视频序列；然后解码器再通过对低分辨率的图像进行上采样从而得到不同分辨率的视频序列。本课题最终实现了AVS可伸缩视频编解码系统。 文章最后对下一步的研究方向进行了简要的讨论。
[Abstract]:With the development of video compression and coding technology and network technology, video based on network transmission has been widely used. For a long time, the development of digital audio and video industry in China has been restricted by foreign video compression coding standards, and it has to pay a high patent fee every year. For this reason, AVS working group has been established in our country, and the digital audio and video codec standard AVS.AVS, which has our own intellectual property rights, has the advantages of high efficiency, low complexity of coding and decoding algorithm and low cost of software and hardware implementation. The future is bound to be widely used in the field of digital audio and video. The video coding based on network transmission is very different from that of storage oriented video coding. It must take into account the heterogeneity of network and the diversity of terminal equipment, so the code stream compiled by encoder should be in space and time. Quality and other aspects can be scalable, such a stream has a good adaptability, scalable video coding (Scalable Video Coding, SVC) technology came into being in such an environment. SVC realized only one video sequence coding, The receiving terminal can decode or transmit part of the coded code stream according to the network connection and its own display size so as to keep the bandwidth balanced and clear and smooth multi-screen display. There is no doubt that the in-depth study of scalable video coding and decoding technology is of great significance. The design goal of this thesis is to realize scalable video codec system, so the most important work is to realize an optimized AVS decoder, and then modify the AVS codec algorithm to realize the scalable coding in AVS space and the correct decoding of the code stream. The main contents and results are as follows: based on the in-depth study of AVS codec principle and official reference decoding software, a new decoding process is proposed. At first, a new AVS decoder is implemented with C code on PC computer, and the decoder is optimized at C level. The MMX instruction set and the SSE2 instruction set in SIMD technology are used to rewrite the code segments which are computationally concentrated and time-consuming in the new decoder. The optimized new decoder realizes the real-time decoding of D 1 image based on PC platform. By studying the principle and key technologies of scalable video coding, a scalable video coding and decoding scheme in AVS space is proposed, which is simpler than SVC. Firstly, the encoder is modified according to the scheme. The modified code stream can provide a variety of video sequences with different resolution. Then the decoder gets the video sequence with different resolution by up-sampling the low resolution image. Finally, the AVS scalable video coding and decoding system is implemented. Finally, the next research direction is briefly discussed.
【学位授予单位】：太原理工大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TN919.81
本文编号：2397133
[Abstract]:With the development of video compression and coding technology and network technology, video based on network transmission has been widely used. For a long time, the development of digital audio and video industry in China has been restricted by foreign video compression coding standards, and it has to pay a high patent fee every year. For this reason, AVS working group has been established in our country, and the digital audio and video codec standard AVS.AVS, which has our own intellectual property rights, has the advantages of high efficiency, low complexity of coding and decoding algorithm and low cost of software and hardware implementation. The future is bound to be widely used in the field of digital audio and video. The video coding based on network transmission is very different from that of storage oriented video coding. It must take into account the heterogeneity of network and the diversity of terminal equipment, so the code stream compiled by encoder should be in space and time. Quality and other aspects can be scalable, such a stream has a good adaptability, scalable video coding (Scalable Video Coding, SVC) technology came into being in such an environment. SVC realized only one video sequence coding, The receiving terminal can decode or transmit part of the coded code stream according to the network connection and its own display size so as to keep the bandwidth balanced and clear and smooth multi-screen display. There is no doubt that the in-depth study of scalable video coding and decoding technology is of great significance. The design goal of this thesis is to realize scalable video codec system, so the most important work is to realize an optimized AVS decoder, and then modify the AVS codec algorithm to realize the scalable coding in AVS space and the correct decoding of the code stream. The main contents and results are as follows: based on the in-depth study of AVS codec principle and official reference decoding software, a new decoding process is proposed. At first, a new AVS decoder is implemented with C code on PC computer, and the decoder is optimized at C level. The MMX instruction set and the SSE2 instruction set in SIMD technology are used to rewrite the code segments which are computationally concentrated and time-consuming in the new decoder. The optimized new decoder realizes the real-time decoding of D 1 image based on PC platform. By studying the principle and key technologies of scalable video coding, a scalable video coding and decoding scheme in AVS space is proposed, which is simpler than SVC. Firstly, the encoder is modified according to the scheme. The modified code stream can provide a variety of video sequences with different resolution. Then the decoder gets the video sequence with different resolution by up-sampling the low resolution image. Finally, the AVS scalable video coding and decoding system is implemented. Finally, the next research direction is briefly discussed.
【学位授予单位】：太原理工大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TN919.81
【参考文献】
相关期刊论文 前3条
1 杨少博;李凤亭;;AVS自适应环路滤波器硬件设计与实现[J];电视技术;2008年06期
2 谢黎;张睿;;基于XSBase270平台的AVS解码器亚像素插值优化[J];计算机应用与软件;2010年07期
3 高文;王强;马思伟;;AVS数字音视频编解码标准[J];中兴通讯技术;2006年03期
，本文编号：2397133
本文链接：https://www.wllwen.com/falvlunwen/zhishichanquanfa/2397133.html
