视频在传输过程中会因为信道噪声等原因导致失真,采取差错控制方可以保证视频传输的质量。传统的差错控制方法将视频当成普通的比特流,而忽视了视频的内容特征。提出的基于内容的差错控制方法按照内容特征划分视频结构单元并分配相应的差错控制方法。实验结果表明,该差错恢复方法不仅能有效利用信道带宽,而且可保护显著性区域,使重建的视频更加符合用户的视觉需求,实现用小的代价保护重要的视频内容。
Video signals may be distorted during communication due to channel noises. Error control methods are needed to ensure the quality. Traditional schemes treat video as ordinary bit streams, ignoring the underlying content of the video. A content-aware error control scheme proposed in this paper uses content feature to classify the content of video into a set of unit classes, and associate each unit class with representative error control approaches. Experiment results show that using this error control schemes during error resilience process, the bandwidth canbe utilized efficiently, and Focus Of Attention (FOA) can be protected, the reconstructed video can satisfy the user's demand of information maximally at the cost of minim resources.