Current Issue Cover
视频压缩编码的差错复原技术

梁 凡1, 魏晓晖1, 肖自美1, 刘红梅1(中山大学电子与通信工程系,广州 510275)

摘 要
在无线移动信道和因特网等环境下的压缩视频传输业务,信道传输差错不但严重影响业务质量,甚至会导致整个视频通信完全失效,因此差错复原技术就成为易发生差错信道下视频编码的重要组成部分.在H.263+,H.263++和MPEG-4等视频编码标准中均采用了若干差错复原技术或工具.该文对目前各种新的视频编码标准中所应用的一些差错复原技术,如重同步、数据分割、可逆变长编码、参考图象选择、差错掩盖等的内容、原理进行了详细的叙述.并利用第三代移动通信系统的信道模型进行了差错复原技术的实验.结果表明,在应用了差错复原技术后重建图象的主客观质量均得到了明显的改善.
关键词
Error Resilience in Video Coding

()

Abstract
The problem of error resilience in compressed video communication is becoming increasingly important because of the growing interest in video delivery over error-prone channels such as mobile networks and the Internet. One inherent problem of such error-prone channels is channel errors. Current video compression standards achieve efficient compression by using predictive coding and variable length entropy coding. Variable length coding schemes are highly vulnerable to channel errors. Predictive coding make matters much worse since the errors quickly propagate across the entire video sequence. So the effect of these errors not only can cause great degradation of the quality of service, even can lead to the failure of the whole video communication system. Some error resilience technologies have already been included in several video coding international standards such as H.263+, H.263++and MPEG-4. This article gives a detail description of the content and principle of error resilience technologies, such as resynchronization, data partitioning, reversible variable length coding, reference picture selection mode and error concealment, etc, which are used in diversified new video coding international standards. This article also gives some simulation results of error resilience technologies based on IMT-2000 channel models. The result shows error resilience technologies can remarkably improve the subjective and objective quality of the reconstructed image.
Keywords

订阅号|日报