摘要
针对不可靠传输会导致视频码流质量下降的问题,提出了一种基于编码域的自适应灵活宏块次序(FMO)方法用于视频容错编码。首先提取出图像的运动区域,然后依据对传输环境和图像特征的分析,计算出该运动区域的错误敏感度。在错误敏感度的基础上,计算最优的FMO编码模式,以使容错强度能够更有自适应性。为提高计算效率,主要的运算信息均来自编码域,即编码过程中产生的结果。实验结果表明:相较于固定FMO模式的方法,文章提出的方法能够更为显著地改善不可靠传输时的视频质量,且计算效率较高。
Coded video bitstream suffers significant quality degradation when transmitted over unreliable routes. In this paper, an adaptive flexible macroblock ordering （FMO） scheme based on encoding domain for error-resilient encoding is proposed to cope with this challenge. A motion area extraction method is first utilized, and the error sensitivity of motion area is then computed according to both transmission condition and input picture features. Based on the error sensitivity, optimal FMO encoding mode is calculated, which makes the error-resilient intensity more adaptive. To lower the complexity, most of the information for computations is from the encoding domain, i.e. the result produced in encoding process. Experiments prove that the proposed scheme outperforms fixed-pattern schemes and improves the video quality noticeably in a very efficient way.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2011年第9期2127-2131,共5页
Chinese Journal of Scientific Instrument
关键词
灵活宏块次序
编码域
错误敏感度
运动区域提取
容错
视频编码
H.264
flexible macroblock ordering （FMO）
encoding domain
error sensitivity
motion area extraction
error resilience
video encoding
H. 264
