Abstract: Network-based audio and video calls as well as video conferencing may loss packets because of limited network bandwidth, which results in video streaming quality problems and reduces the effects of video calls and video conferencing. The real-time streaming quality control algorithm was proposed. This algorithm adopted time control approach to detect and remove the bad key frames, thus reducing the occurrence of blurred screen. The proposed algorithm can efficiently reduce the costs of time and space, which eventually enhances the streaming fluency. The experiments were conducted in terms of original frame playback, post-processing playback, and key frame pretreatment playback. The experimental comparisons showed that the proposed algorithm could efficiently improve the quality and fluency of playback, moreover, the computational complexity of playing processing was also decreased by more than 40%. The results show that the proposed algorithm has outstanding effect to improve the quality of live streaming and reduce the occurrence of blurred screen.

Key words: H.264, real-time streaming media, key frame preprocessing, timing control algorithm, Group Of Pictures (GOP)

摘要： 网络带宽特别是上行带宽受限情况下,基于H.264高压缩率的网络音视频通话或视频会议普遍会出现不同程度的丢包现象,从而造成流媒体播放的花屏等质量问题,影响视频通话或视频会议的效果。针对上述问题,提出了基于关键帧预处理的实时流媒体播放质量控制方法。该方法采用单元时序控制算法对视频关键坏帧进行实时侦测和取舍,进而减少花屏情况。使用该方法将降低后处理计算的时空成本,并提高流媒体播放的流畅性。通过原始帧播放、后处理播放、关键帧预处理播放3种处理方法的实验对比,证明基于播放单元时序控制算法的实时流媒体播放器,不但明显提高了播放的流畅性,播放后处理的计算复杂度也降低了40%以上。结果表明,该方法对提高播放质量、减少花屏情况有着突出的效果。

关键词: H.264, 实时流媒体, 关键帧预处理, 时序控制算法, 画面组单元