基于内容丢帧的视频自适应传输优化策略

摘要

According to user’s preference to video content, a measurement for importance of video content with semantic event and exciting intensity is proposed. A multiple granularity of semantic content based frame dropping model in temporal domain is constructed. The video adaptive transmission strategy based on semantic frame dropping is designed. Results of the experiments indicate that the proposed strategy achieves much more improvement in network performance and semantic quality than traditional approach of frame-type based frame dropping.%根据用户对视频内容的个性化偏好，提出结合语义事件与精彩度的视频内容分级方法，构建时域内基于多粒度语义内容的统一丢帧模型，在 RTP/RTSP 流媒体服务器中基于语义丢帧算法设计了视频自适应传输优化策略。实验结果表明，本文提出的基于内容的丢帧策略在网络性能、语义质量等方面优于传统基于帧类型的丢帧策略。