基于WMSNs的视频编码技术研究
发布时间：2018-07-27 12:36
【摘要】：随着视频、音频等复杂信息在人们生活中各个领域的广泛应用,以无线网络通信技术和传感器技术为核心理论的无线多媒体传感器网络(WMSNs)诞生了。WMSNs主要由多媒体视频音频生成设备以及其它特定作用的传感器组成。它在实际生活中的主要应用是对目标环境中数据量庞大的复杂多媒体信息进行采集、编解码处理和无线传输等。由于WMSNs的存储容量和计算能力等资源受到网络限制,所以研究其低复杂度、高压缩效率的视频信息编解码技术成为当前研究的主要问题。本文对WMSNs中多媒体视频信息编解码技术的有效方法进行了进一步的理论研究。介绍了两部分理论知识:压缩感知(CS)和分布式视频编码算法(DVC),并全面地分析了分布式压缩感知(DCS)视频编解码方法。本文重点对CS视频压缩技术中稀疏变换算法进行了改进,提出了联合稀疏变换(JST)方法和自适应双分组稀疏变换(ADGST)方法。JST方法充分利用了小波稀疏变换和冗余字典稀疏变换(RDST)的优点,优化了视频序列的稀疏表示,降低了视频序列编解码时间。ADGST方法是在自适应组稀疏变换(AGST)基础上对次要帧稀疏变换进行的优化,先将次要帧进行小波稀疏变换,再对小波稀疏变换后的相似图像进行自适应分组,对同组内的相似图像进行相同的RDST变换,减少了视频序列编码时间。
[Abstract]:With the wide application of video, audio and other complex information in every field of people's life, The wireless multimedia sensor network (WMSNs), which is based on wireless network communication technology and sensor technology, is mainly composed of multimedia video and audio generating equipment and other special sensors. Its main application in real life is to collect, encode and decode the complicated multimedia information in the target environment. Because the storage capacity and computing power of WMSNs are limited by the network, the research of video coding and decoding technology with low complexity and high compression efficiency has become the main problem. In this paper, the effective methods of multimedia video coding and decoding in WMSNs are further studied. This paper introduces two parts of theoretical knowledge: compression aware (CS) and distributed video coding algorithm (DVC), and analyzes the distributed compression aware (DCS) video coding and decoding methods. This paper focuses on the improvement of sparse transform algorithm in CS video compression technology. In this paper, the joint sparse transform (JST) method and the adaptive double packet sparse transform (ADGST) method are proposed to optimize the sparse representation of video sequences by taking full advantage of the advantages of wavelet sparse transform and redundant dictionary sparse transform (RDST). The method of reducing the encoding and decoding time of video sequence. ADGST is an optimization of sparse transform of secondary frames based on adaptive group sparse transform (AGST). Then the similar image after sparse wavelet transform is self-adaptively grouped, and the similar image in the same group is transformed by the same RDST transform, which reduces the coding time of video sequence.
【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
本文编号：2147849
[Abstract]:With the wide application of video, audio and other complex information in every field of people's life, The wireless multimedia sensor network (WMSNs), which is based on wireless network communication technology and sensor technology, is mainly composed of multimedia video and audio generating equipment and other special sensors. Its main application in real life is to collect, encode and decode the complicated multimedia information in the target environment. Because the storage capacity and computing power of WMSNs are limited by the network, the research of video coding and decoding technology with low complexity and high compression efficiency has become the main problem. In this paper, the effective methods of multimedia video coding and decoding in WMSNs are further studied. This paper introduces two parts of theoretical knowledge: compression aware (CS) and distributed video coding algorithm (DVC), and analyzes the distributed compression aware (DCS) video coding and decoding methods. This paper focuses on the improvement of sparse transform algorithm in CS video compression technology. In this paper, the joint sparse transform (JST) method and the adaptive double packet sparse transform (ADGST) method are proposed to optimize the sparse representation of video sequences by taking full advantage of the advantages of wavelet sparse transform and redundant dictionary sparse transform (RDST). The method of reducing the encoding and decoding time of video sequence. ADGST is an optimization of sparse transform of secondary frames based on adaptive group sparse transform (AGST). Then the similar image after sparse wavelet transform is self-adaptively grouped, and the similar image in the same group is transformed by the same RDST transform, which reduces the coding time of video sequence.
【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
【参考文献】
相关期刊论文 前5条
1 杨冰;邓曙光;文双春;;无线多媒体传感器网络能量均衡的QoS路由算法[J];传感器与微系统;2014年04期
2 陈长英;杨秀红;付勇;郑晓势;杜龙安;;无线多媒体传感器网络的关键技术研究进展[J];山东科学;2013年03期
3 沙超;王汝传;黄海平;孙力娟;;一种基于多目标遗传优化的无线多媒体传感器网络节能覆盖方法[J];电子学报;2012年01期
4 周灵;王建新;;无线多媒体传感器网络路由协议研究[J];电子学报;2011年01期
5 杨卓静;孙宏志;任晨虹;;无线传感器网络应用技术综述[J];中国科技信息;2010年13期
相关博士学位论文 前1条
1 王镇道;视频压缩的运动估计与小波方法研究[D];湖南大学;2008年
相关硕士学位论文 前3条
1 丁雨廷;LDPC码编译码原理及多码率编码研究[D];杭州电子科技大学;2014年
2 黄涛;基于压缩感知的分布式视频编解码研究[D];浙江师范大学;2012年
3 于新波;视频压缩中的预处理及运动估计算法的研究[D];山东大学;2008年
，本文编号：2147849
本文链接：https://www.wllwen.com/kejilunwen/wltx/2147849.html
