首页 | 本学科首页   官方微博 | 高级检索  
     

一种基于感兴趣区域的质量可分级编码传输方案研究
引用本文:宋晓炜,刘忠,杨蕾,吴源昭.一种基于感兴趣区域的质量可分级编码传输方案研究[J].中原工学院学报,2013(2):49-53.
作者姓名:宋晓炜  刘忠  杨蕾  吴源昭
作者单位:中原工学院,郑州450007
基金项目:国家科技重大专项项目(2009ZX03004-006); 国家自然科学基金项目(60902063); 中国博士后科学基金资助项目(20090460696); 河南省高等学校青年骨干教师资助计划项目(2010GGJS-098); 河南省科技创新人才支持计划项目(124100510015)
摘    要:采用SVC中的MGS质量可分级编码技术,提出了一种基于感兴趣区域的质量可分级编码传输方案.首先,将图像区域划分为感兴趣区域与背景区域两部分,然后,采用MGS质量可分级编码技术对感兴趣区域的图像进行逐级增强,对残差(增强层减去基本层)的16个DCT变换系数进行分层传输得到不同质量的增强层,对于背景区域,抑制增强层的传输,只传输基本层视频图像.结果表明,此方法可以在网络带宽受限或者传输码率下降的情况下,通过抑制背景区域增强层的传输来进一步降低码率,同时保证了感兴趣区域图像的可分级编码质量.

关 键 词:MGS  感兴趣区域  质量可分级编码  区域分级传输  选择性增强

A Transmitting Method for Quality Scalability Coding Based on Region of Interest
SONG Xiao-wei,LIU Zhong,YANG Lei,WU Yuan-zhao.A Transmitting Method for Quality Scalability Coding Based on Region of Interest[J].Journal of Zhongyuan Institute of Technology,2013(2):49-53.
Authors:SONG Xiao-wei  LIU Zhong  YANG Lei  WU Yuan-zhao
Affiliation:(Zhongyuan University of Technology, Zhengzhou 450007, China)
Abstract:In this paper, a transmitting method based on region of interest has been proposed for quality scalability video coding by using MGS technology in SVC. Firstly, video images are artificially segmented into two partitions, ROI and background. Subsequently, for ROI, the 16 DCT coefficients of the residual, which are acquired by subtracting base layer from enhance layer, are transmitted hierarchically to the enhance layer. By employing the quality scalability coding technology of MGS, the quality of ROI is gradually enhanced. For the background, the 16 DCT coefficients are controlled not to transmit to the enhance layer, resulting that only base layer is assigned to the region. Experimental re- sults show that coding bitrate can be further reduced without affecting the quality of ROI, for which the adaptability of the network to the limitation of its bandwidth is improved.
Keywords:MGS  region of interest  quality scalability coding  region scalability transmitting  selective enhancement  
本文献已被 CNKI 维普 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号