多视点视频无线软传输技术研究  

MULTI-VIEW VIDEO SOFTCAST OVER WIRELESS CHANNEL

在线阅读下载全文

作  者:桑涛 唐述 Sang Tao;Tang Shu(IT Management Department,China Metallurgical Group Corporation,Beijing 100028,China;College of Computer Science and Technology,Chongqing University of Posts and Telecommunications,Chongqing 400065,China)

机构地区:[1]中国冶金科工集团有限公司信息化管理部,北京100028 [2]重庆邮电大学计算机科学与技术学院,重庆400065

出  处:《计算机应用与软件》2022年第10期126-131,161,共7页Computer Applications and Software

基  金:国家自然科学基金项目(61601070)。

摘  要:由于三维视频包含从不同角度同时拍摄的多视角视频导致数据量成倍增加。此外,基于独立信源信道编码的传统视频传输策略由于无线信道的时变特性并不能很好地适应于移动场景。近来,一种被称为软传输的非编码传输方案被提出用于高效的二维视频无线传输。然而,该方案并不能实现高效的三维或多视点视频传输。针对多视点视频提出一种优化的软传输方案。为了降低多视点间冗余,线性变换扩展为四维离散余弦变换;基于离散余弦变换系数分布特性,对辅助数据的传输进行优化进一步降低传输块平均能量。实验结果表明,该方法可实现多视点视频传输性能5~8 dB的提升。Due to the fact that 3D video(3DV)contains multi-view video(MVV)simultaneously captured from different perspectives,the data amount is multiplied.The conventional video transmission scheme based on separate source coding and channel coding may fail to the broadcast and mobile scenarios due to the drastic changes in wireless channel condition.Recently,an uncoded transmission scheme called SoftCast has been presented for efficient 2D video wireless transmission.However,it cannot be directly applied to 3DV or MVV wireless transmission with high ef?ciency.This paper proposes an optimized SoftCast scheme for MVV transmission.In order to eliminate the inter-view redundancy among multiple viewpoints,the linear transform was extended to four-dimensional discrete cosine transform(DCT).Based on the rapid decaying property of DCT distribution,the metadata transmission was optimized to further reduce the mean energy of chunks.The experimental results show that the proposed method can improve the multi-view video transmission performance by 5 dB~8 dB.

关 键 词:软传输 多视点视频 离散余弦变换 三维视频 

分 类 号:TP37[自动化与计算机技术—计算机系统结构] TN914.53[自动化与计算机技术—计算机科学与技术]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象